15558/6/06

JULY 2006

SURFACE INSPECTION
METHODS FOR DETECTION
OF STRESS CORROSION
CRACKING IN AUSTENITIC
AND DUPLEX STAINLESS
STEELS: FINAL REPORT

For: A Group of Sponsors

CONTENTS

EXECUTIVE SUMMARY i
Background i
Objective i
Experimental Approach i
Results and Discussion ii
Conclusions iii
Recommendations iv
Financial Statement iv

1. INTRODUCTION 1

2. OBJECTIVE 1

3. EXPERIMENTAL PROCEDURE 2
3.1. TEST SAMPLE PREPARATION 2
3.1.1. Materials 2
3.1.2. Hot Tensile Tests 2
3.1.3. Plain Plates 2
3.1.4. Welded Plates 2
3.1.5. Welded Pipes 2
3.1.6. Materials Characterisation 2
3.2. DROP EVAPORATION TESTS 3
3.2.1. Tests on Plain and Welded Plates 3
3.2.2. Tests on Welded Pipes 4
3.3. NDE METHODS 5
3.3.1. General Procedure 5
3.3.2. Penetrant Testing (PT), Magnetic Particle Inspection (MPI) and
Alternating Current Potential Drop (ACPD) 5
3.3.2.1. Description of equipment used 5
3.3.2.2. Procedure (plates and pipes) 5
3.3.2.3. 316L stainless steel welded plates 6
3.3.3. Acoustic Emission (AE) 6
3.3.4. Comparative Vacuum Monitoring (CVM) 6
3.3.5. Eddy Current Testing 6
3.3.5.1. Scope 6
3.3.5.2. Description of equipment used 7
3.3.5.3. Procedure 7
3.3.5.4. Variation of performance with lift off 7
3.3.5.5. Adaptation for high temperatures 7
3.3.6. Alternating Current Field Measurement (ACFM) 8
3.3.6.1. Scope 8
3.3.6.2. Description of equipment used 8
3.3.6.3. Procedure 8
3.3.6.4. Adaptation to hot testing 8
3.3.7. Ultrasonic Testing 8
3.3.7.1. Scope 8
3.3.7.2. Description of equipment used 9

15558/6/06
Copyright © 2006, TWI Ltd
3.3.7.3. Procedure 9
3.3.8. Eddy Current Array 9
3.3.8.1. Scope 9
3.3.8.2. Description of equipment used 9
3.3.8.3. Procedure 9
3.3.9. Pulsed Eddy Current 9
3.3.9.1. Description of equipment used 9
3.3.9.2. Procedure 9
3.3.10. EMAT and Laser Ultrasound Tests 10
3.3.10.1. Scope 10
3.3.10.2. Description of equipment used 10
3.3.10.3. Procedure 10
3.3.11. Thermal Methods 10
3.3.12. Laser Shearography 10

4. RESULTS 10
4.1. HOT TENSILE TESTS 10
4.2. MECHANICAL LOADING AND DROP EVAPORATION TESTS 11
4.2.1. Austenitic Stainless Steel Plates 11
4.2.1.1. Loading and drop evaluation 11
4.2.1.2. Dye penetrant inspection 11
4.2.2. Duplex Stainless Steel Plates 11
4.2.2.1. Loading and drop evaporation 11
4.2.2.2. Penetrant inspection 12
4.2.3. Welded Duplex Stainless Steel Pipes 12
4.2.3.1. Loading and drop evaporation 12
4.2.3.2. Dye penetrant inspection 12
4.3. METALLOGRAPHY AND FRACTOGRAPHY 12
4.4. NDE METHODS 13
4.4.1. General 13
4.4.2. Dye Penetrant Inspection /MPI/ACPD 13
4.4.3. Acoustic Emission 13
4.4.4. Comparative Vacuum Monitoring 14
4.4.5. Eddy Current Inspection 14
4.4.6. Alternating Current Field Measurement (ACFM) 14
4.4.7. Ultrasonic Testing (UT) 14
4.4.8. Eddy Current Array 15
4.4.9. Pulsed Eddy Current 15
4.4.10. EMAT and Laser Ultrasound Tests 15
4.4.11. Thermal Methods 15
4.4.12. Laser Shearography 16

5. DISCUSSION 16
5.1. PRODUCTION OF STRESS CORROSION CRACKS 16
5.2. EFFECTIVENESS OF NDE METHODS IN DETECTING SCC 17

6. CONCLUSIONS 19

15558/6/06
Copyright © 2006, TWI Ltd
7. RECOMMENDATIONS 20

8. ACKNOWLEDGEMENTS 21

9. REFERENCES 21

TABLES 1-8
FIGURES 1-61

APPENDIX A - WELD SPECIFICATIONS

APPENDIX B - RESULTS FROM THE DP, MPI and ACPD of Cracks

APPENDIX C - ACOUSTIC EMISSION TESTS

APPENDIX D - EDDY CURRENT TESTS

APPENDIX E - ACFM TESTS

APPENDIX F - ULTRASONIC PHASED ARRAY RESULTS

APPENDIX G - PULSED EDDY CURRENT TEST REPORT

APPENDIX H - EMAT AND LASER ULTRASOUND TEST REPORT

APPENDIX I - SHEAROGRAPHY TEST REPORT

15558/6/06
Copyright © 2006, TWI Ltd
EXECUTIVE SUMMARY
Background
Chloride induced stress corrosion cracking (SCC) of austenitic stainless steels and duplex
stainless steels (DSS) can cause rapid failure once initiated. This environmentally assisted
cracking mechanism leads to failure of the material at a stress less than the ultimate tensile
stress (UTS) and in conditions where corrosion alone would not necessarily lead to failure.
The presence of residual stresses at welds increases the likelihood of SCC occurring at such
locations when the equipment is subject to the necessary environmental conditions. These are
elevated temperature, chloride concentration and appropriate potential, which is typically
controlled by oxygen availability. In the case of DSS, SCC has been observed in the apparent
absence of oxygen at extreme conditions of temperature and chloride concentration.
Operating temperatures may be as high as 130°C for duplex stainless steels, or possibly
higher for superduplex stainless steels and there have been instances of recent failures of
topside equipment under such conditions. Austenitic stainless steels are susceptible to SCC at
lower temperatures, typically above 55°C for 300-series alloys.

For the purposes of detecting flaws in structures, many non-destructive test methods are
available and in most cases the optimum method can be chosen from a range of techniques.
However, chloride stress corrosion cracking in austenitic and duplex stainless steels is
difficult to assess because the crack morphology typically comprises very fine, branching
cracks, which do not produce a response from most NDT systems.

There is a need for an improved method(s) which can be applied whilst the plant/facility is
operational at temperature, allowing more frequent inspection, without leading to expensive
downtime nor compromising safety for personnel, plant and environment. Methods are also
required for the detection of SCC during plant shutdowns, when the cracks might be under
compression and harder to find using conventional methods. This project assesses a number
of techniques that may be suitable for application to operational plant and equipment.

Objective
To assess relative performance of established, recently developed and novel inspection
methods to detect chloride-induced SCC in austenitic and duplex stainless steels.

Experimental Approach
In order to assess NDE methods, material first had to be exposed to environmental/loading
conditions that would, within a reasonable period, result in SCC. Initial tests were conducted
using plain and welded plates (340mm x 150mm x 12mm). These were stressed in three-point
bend test jigs designed and manufactured for this project. Two 22Cr DSS girth-welded pipes
were also produced and loaded, intact, within in a full pipe four-point bend rig. The materials
were heated to an appropriate temperature (70°C for type 316L and 150°C for 22Cr DSS
plates and pipes) to allow SCC to initiate and natural seawater dripped onto the surface.

Loading of the samples was applied after the samples reached the required temperatures.
Strain was monitored using attached strain gauges and increased to achieve a maximum stress
in excess of the 0.2% proof stress of the test material. Following mechanical loading, natural
seawater was continuously dripped on to the top surface of the plates and pipes at several

15558/6/06
Copyright © 2006, TWI Ltd
locations aligned along the transverse centreline of the plates or across the weld, as
appropriate. The dripping rate was 0.24 ml/min and remained constant for the duration of the
tests.

The plates were periodically checked to establish the onset of SCC using hot dye penetrant
testing. A range of NDE techniques were evaluated, including comparative vacuum
monitoring and acoustic emission. Evaluation of other NDE techniques, after cracks have
formed, included Magnetic Particle Inspection (MPI), eddy current testing, Alternating
Current Potential Difference (ACPD), Alternating Current Field Measurement (ACFM),
phased array ultrasonics, eddy current arrays, pulsed eddy currents, surface waves induced by
lasers and electromagnetic acoustic transducers (EMATS), thermography and laser
shearography.

Results and Discussion

Dye penetrant inspection was successful in detecting SCC in both materials at high
temperatures and magnetic particle inspection was successful in the duplex stainless steel at
room temperature. The comparative vacuum method was initially affected by the mechanical
effects of the salt deposits that resulted in some extra leaks and blockages, but one set of
cracks was successfully detected, although the cracking was quite extensive before detection
took place.

Acoustic emission achieved some success in detecting crack propagation in the duplex pipe
samples. This contrasted with poor success rates in detecting SCC in the austenitic stainless
steel plate samples. However, with the two duplex pipe tests, there was considerable variation
in the acoustic response. As regards field use, optimisation of the threshold values to flag
significant acoustic emission signals is important for the success of this method. Additionally
the success of acoustic emission is highly dependent on the site conditions (e.g. the level and
nature of background noise and acoustic signal attenuation) and complex geometry (e.g.
branch connection). Even with a carefully optimised monitoring system, effective site
monitoring of SCC in duplex stainless steel in process plant has to be a compromise between
detection confidence levels and false calls, i.e. high levels of detection accompanied by large
levels of spurious results or lower levels of detection accompanied by lower levels of false
calls. In the present work, location of the acoustic emission sources was relatively accurate.
However, the sample geometries were relatively simple. More complex geometries will add
to additional complexity in source location, which might need to be addressed by computer
modelling of the geometry concerned.

The eddy current high frequency methods and phased array ultrasonics appeared to have
similar detection capabilities at low temperatures on the 316L plates. On the duplex plates the
electromagnetic methods (eddy current and ACFM) could be adapted for high temperatures.
This however meant some loss in sensitivity with the adaptations used.

The MPI was capable of giving good indications of the presence of cracks at room
temperature for the duplex material. ACPD gave an indication of severity, but undersized the
pipe cracks.

The EMAT Rayleigh wave system has potential for application at high temperatures but its
sensitivity to small cracks was not tested and it is still a research tool, as is the pulsed eddy

ii

15558/6/06
Copyright © 2006, TWI Ltd
current method, which did not appear to have any particular advantages over the other eddy
current methods. The laser shearography method could be adapted to high temperature
application and was able to detect cracks as small as a 3mm deep ACPD indication, but was
impractical for other reasons.

Conclusions
1. Drop evaporation testing of austenitic type 316L and 22Cr duplex stainless steel
plates loaded in three-point bend, and girth-welded pipes loaded in four-point bend
test rigs provided a robust technique for studying SCC. For both 22Cr duplex and
316L materials, stress corrosion cracks always formed in the parent material or
HAZ outside of the welds.

2. The acoustic emission method was able to operate at high temperatures and
successfully detected cracking in the duplex stainless steel pipes. Tests involving
the plate samples were inconclusive, due in part to the small size of the plates.
Monitoring using acoustic emission to detect SCC can either be done by setting low
thresholds at the expense of increasing the number of spurious calls, or by setting
higher thresholds to eliminate spurious calls at the expense of possibly missing
some instances of SCC propagation.

3. The hot dye penetrant method was able to detect cracks in both sample types to a
high sensitivity and at temperature. From the limited evidence available, for
inspecting 316L stainless steel at temperatures up to 60°C, it could be replaced by
eddy currents (single probe or arrays), or phased array ultrasonics.

4. At higher temperatures (up to 150°C) it was found practical to cool the eddy current
weldscan probe and ACFM probes by an airflow. This however decreased the
sensitivity relative to hot dye penetrants. The phased array ultrasonics could not be
deployed at these temperatures.

5. If the electromagnetic or ultrasonic methods are to replace dye penetrant for

inspection of austenitic type 316L or 22Cr duplex stainless steels at low
temperatures then comparison over a wider range of crack sizes is needed. For
application to hot surfaces, the ACFM and eddy current methods used in this project
would need further development. Pulsed eddy current testing offered no advantage
over conventional methods.

6. The EMAT system could detect cracks in plate samples from both the top and
bottom faces by Rayleigh waves, and could possibly be developed for hot
application. The Rayleigh wave method might also offer a method of detecting sub-
surface flaws, due to internal cracking.

7. The MPI method could be used on the DSS specimens for delineating the cracks at
room temperature. This was not possible at high temperatures. ACPD was used
successfully to give an indication of crack severity, but undersized the cracks in the
duplex stainless steel pipe.

iii

15558/6/06
Copyright © 2006, TWI Ltd
8. The comparative vacuum, shearography and thermography methods are not suitable
for the detection of SCC.

Recommendations
For the detection of the propagation of stress corrosion cracks, acoustic emission should be
considered. However, further refinement may be necessary to determine safe alarm
thresholds that allow any significant crack growth to be detected, whilst not triggering an
undue number of false calls.

For stress corrosion cracking at moderate temperatures in austenitic stainless steels, eddy
currents (single probe or arrays), or phased array ultrasonics should be considered as
complementary to hot penetrant testing. However, comparison over a wider range of crack
sizes is needed to better establish real time performance.

At temperatures significantly greater than 60°C (150°C in the present work), hot penetrant
testing proved to offer superior sensitivity to electromagnetic methods (ACFM or eddy
current) using cooled probes. Improved cooling designs could reduce the difference in
sensitivity and should be investigated where use of dye penetrant is inconvenient.

The Rayleigh wave method should be further investigated as a means of detecting sub-
surface flaws in hot materials.

Further work needs to be carried out on a wider range of realistic samples if a statistical level
of confidence of the capability of any technique is required.

Financial Statement
The project had a total funding of £275,000 which has been fully utilised during the course of
the project.

Sponsor Group
The project was supported by the following companies:

- Lloyds Register
- Talisman Energy (UK) Ltd
- Health & Safety Executive
- Shell UK Ltd
- UK Nirex Ltd
- DSM TechnoPartners (first year of study)

iv

15558/6/06
Copyright © 2006, TWI Ltd
1. INTRODUCTION
Chloride induced stress corrosion cracking (SCC) of austenitic stainless steels and duplex
stainless steels (DSS) can cause rapid failure once initiated. This environmentally assisted
cracking mechanism leads to failure of the material at a stress less than the ultimate tensile
stress (UTS) and in conditions where corrosion alone would not necessarily lead to failure.
The presence of residual stresses at welds increases the likelihood of SCC occurring at such
locations when the equipment is subject to the necessary environmental conditions. These are
elevated temperature, chloride concentration and potential, which is typically controlled by
oxygen availability. In the case of DSS, SCC has been reported to occur in the apparent
absence of oxygen at extreme conditions of temperature and chloride concentration.
Operating temperatures may be as high as 130°C for duplex stainless steels, or possibly
higher for superduplex stainless steels and there have been instances of recent failures of
topside equipment under such conditions. Austenitic stainless steels are susceptible to SCC at
lower temperatures, typically above 55°C for 300 series alloys.

Within overall corrosion management systems and schemes, it is necessary to inspect

equipment to demonstrate that this potential damage mode has not occurred when operating
temperatures are above the minimum required for SCC, if water and chlorides are also
present. Presently, no reliable standard NDE method exists which is recognised as suitable
for inspection for SCC whilst the plant is in-service.

To provide industry with the means to manage the risk of failure due to SCC, inspection is
required which will have a high probability of detection of surface cracking whilst the plant is
operational (i.e. hot pipework). Due to the rapid propagation of such cracks, the inspection
method will need to be able to detect cracks of relatively narrow width (e.g. 50-100µm). Data
from such an inspection technique will allow appropriate additional measures such as detailed
inspection and remediation to be undertaken.

For the purposes of detecting flaws in structures, many non-destructive test methods are
available and in most cases the optimum method can be chosen from a range of techniques.
However, chloride stress corrosion cracking in austenitic and duplex stainless steels is
difficult to assess because the crack morphology typically comprises very fine, branching
cracks, which do not produce a response from most NDT systems.

There is a need for an improved method(s) which can be applied whilst the plant/facility is
operational at temperature, allowing more frequent inspection, without leading to expensive
downtime nor compromising safety for personnel, plant and environment. Methods are also
required for the detection of SCC during plant shutdowns, when the cracks might be under
compression and harder to find using conventional methods. The work detailed in this report
assesses a number of techniques that were considered to have potential for application to
operational plant and equipment, including established and novel methods. The original
proposal document and Minutes of the Launch and Inaugural meetings are listed in references
1-5.

2. OBJECTIVE
To assess relative performance of established, recently developed and novel inspection
methods to detect chloride-induced SCC in austenitic and duplex stainless steels.

15558/6/06
Copyright © 2006, TWI Ltd
3. EXPERIMENTAL PROCEDURE
3.1. TEST SAMPLE PREPARATION
3.1.1. Materials
Type UNS S31603 (316L) austenitic stainless steel and UNS S32205 duplex stainless steel
(DSS) plates of 12mm thickness were obtained. UNS S32205 duplex stainless steel pipes
were also employed in the latter phase of the programme. The chemical analyses of the plate
and pipe materials are shown in Table 1.

3.1.2. Hot Tensile Tests

Triplicate hot tensile tests were carried out to determine the mechanical properties of 316L
and 22Cr duplex stainless steels at 70°C and 150°C respectively. The stress/strain curves are
shown in Fig.1. This data was employed to convert stress into equivalent strain when loading
the test samples, monitoring via strain gauges.

3.1.3. Plain Plates

Five plates were prepared from duplex stainless steel (referenced as 2205-02, 2205-03 and
2205-04, 2205-05 and 2205-06) and two plates from austenitic stainless steel type 316L
(referenced as 316L-01 and 316L-02). Cracks were induced in these specimens using the
drop evaporation method developed by TWI, as described in section 3.2.1. These plates were
evaluated using the NDE techniques described in section 3.3.

3.1.4. Welded Plates

Full penetration TIG butt welds were produced in 316L stainless steel plates of 12mm
thickness with a J type preparation. Welding of 316L type plates was completed using
matching composition filler metal. Two welded plates, 316L-W1-1 and 316LW1-2, were
prepared for drop evaporation testing. The welding records are shown in Appendix A.

Cracks were induced in these using the drop evaporation method developed by TWI, as
described in section 3.2.1. These plates were evaluated using the NDE techniques described
in section 3.3.

3.1.5. Welded Pipes

22Cr duplex girth-welded pipes, designated as Pipe 1 and Pipe 2, were welded by manual
TIG welding (employing a Zeron100X superduplex filler wire, heat input between 0.8-
1.0kJ/mm and interpass temperatures below 150°C). The welding records are shown in
Appendix A.

Cracks were induced in these using the drop evaporation method, as described in section
3.2.2, and were evaluated using the NDE techniques described in section 3.3.

3.1.6. Materials Characterisation

Following evaluation of tested plates and pipes with NDE, metallography and fractography
were carried out to evaluate crack geometry, local microstructure and nature of the crack

15558/6/06
Copyright © 2006, TWI Ltd
surfaces. Metallographic sections met1, met2, met3, met4 and met5 were prepared from
plates 316L-01, 316L-W1-1, 2205-04 and Pipes 1 and 2, respectively, to define a relationship
between the character of cracking with the range of NDE responses obtained. Following
examination of these sections it was decided that additional sections were required to obtain
sufficient data. Additional metallographic sections were prepared from plate 2205-05 (met6,
met7 and met8), pipes 1 and 2 and plate 316-01 (sections met9, met10 and met11,
respectively). The sections were prepared from locations where crack indications were found
using dye penetrant inspection (DPI) or Eddy current tests. Cracks in plate 316-01, 2205-04
and Pipe 1 and 2 were also broken open in liquid nitrogen to allow examination of the cracks.
The exposed fracture faces were cleaned using warm water to remove the salt deposits and
they were examined at high magnification in a scanning electron microscope (SEM),
allowing salient features to be characterised.

Metallographic specimens were prepared using standard metallographic techniques. The

microstructures of duplex and austenitic stainless steel materials were revealed using an
electrolytic 40%KOH etch and an electrolytic etch with 20%H2SO4 and 0.1g/l NH4CNS,
respectively. These metallographic specimens were examined using light and scanning
electron microscopy in the as-polished and etched conditions. The ferrite volume fraction was
quantified by manual point counting in accordance with the procedure described in ASTM
E562-02 and TWI Recommended Practice 5632/18/93. Grain size measurements were carried
out using the procedure described in ASTM E112-96.

3.2. DROP EVAPORATION TESTS

3.2.1. Tests on Plain and Welded Plates
Testing was carried out by loading plates, 340mm x 150mm x 12mm, in three-point bend test
jigs with local heating and a controlled saline fluid delivery rate. Two three-point bend jigs
were built using 316L type stainless steel. Figure 2 shows a drawing of the three-point bend
jigs that were designed and built for the tests, while Figure 3 shows loaded plates prior to
commencing the drop evaporation tests.

The plates were heated to the required test temperature (70°C for type 316L plain plates and
welds and 150°C for 22Cr DSS) to allow SCC to initiate. Heating of 316L plates was carried
out by conduction using two electric heaters located under the specimen, either side of the
central loading point roller. The welded plates were mounted with the weld root positioned
immediately above the roller while the weld axis was parallel to the roller axis. The
temperature was monitored using a thermocouple that was attached to the top surface of the
plates, directly above the central roller where maximum strain was applied.

Heating of the duplex stainless steel plates was carried out by modifying the above system,
via the addition of a constant stream of hot air under the plate at a position immediately
adjacent to the central roller. This modification allowed the central part of the plate to reach
150°C while maintaining the rest of the plate at a relatively lower temperature. This
facilitated attachment of the acoustic emission and vacuum sensors, as the former was limited
to a maximum operating temperature of approximately 140°C.

After the plates reached the required temperature they were subjected to mechanical stress.
Mechanical load was increased until the strain on the top surface of the plate exceeded that

15558/6/06
Copyright © 2006, TWI Ltd
which corresponded to the 0.2% proof stress of the plate material, as had been derived from
the hot tensile tests (Fig.1). Strain was monitored continuously by two strain gauges that were
attached to the top surface of the plate, at positions directly opposite the central roller and
close to the edges of the plate. Following mechanical loading, natural seawater was dripped
on to the top surface of the plates at an array of points that were aligned along the transverse
centreline of the plate, where the maximum strain was recorded (Fig.2). The fluid delivery
rate was 0.24 ml/min and this remained constant for the duration of the tests in this
programme. Table 2 lists the loading conditions that were employed for the drop evaporation
tests of plain and welded plates.

The plates were periodically inspected to establish the onset of cracking using dye penetrant
inspection.

3.2.2. Tests on Welded Pipes

Testing was carried out by loading the welded pipes in to a four-point bend test rig. Figure 4
shows a sketch of the jig that was used in the tests, while Figures 5 and 6 show the loaded
pipes prior to commencing the drop evaporation tests.

Heating of the pipes was carried out by conduction using two electric heating blankets that
were wrapped around the pipes either side of the girth weld. The temperature was monitored
using a set of two thermocouples that were attached to the top surface of the pipe (12 o’clock
position), adjacent to the girth weld, where maximum strain was applied. This facilitated
attachment of the acoustic emission sensors which were limited to a maximum operating
temperature of approximately 140°C. The pipes were heated to 150°C to allow SCC to
initiate.

Prior to heating, the pipes’ response to UT and dye penetrant testing were fingerprinted to
ensure that the girth welds were free from any detectable flaws. Following finger printing,
acoustic emission sensors were attached to the pipes. After the pipes reached the required
temperature, mechanical loading was commenced. Mechanical loading was increased until
the strain on the top the pipe adjacent to the weld exceeded that which corresponded to the
0.2% proof stress of the material at temperature (Fig.1), using the stress/strain relationship
determined for the duplex stainless steel plates. Strain was monitored continuously by two
strain gauges that were attached to the top surface of the pipe, at either side of the girth weld.

The pipes remained loaded, at temperature, for approximately 50 hours prior to starting the
drop evaporation, to enable recording of the background acoustic emission signals. Table 2
lists the loading conditions that were employed for both of the girth-welded pipes, 1 and 2.
Following the 50hour delay, natural seawater was dripped on the top surface of the pipes at
an array of points that were aligned along the weld centreline (Fig.4). The fluid delivery rate
was 0.24 ml/min, similar to what was used for the plates and remained constant for the
duration of these tests. The pipes were periodically checked visually to establish the onset of
cracking.

15558/6/06
Copyright © 2006, TWI Ltd
3.3. NDE METHODS
3.3.1. General Procedure
The NDE methods which have been evaluated on SCC samples produced in the project are
shown in Table 3. More detail of the inspection procedures is given below. These methods
were assessed with the aim of ultimately selecting the most appropriate NDE techniques that
could be used to detect external and internal SCC in welds at elevated temperature.

It was initially planned to assess these techniques in the test rig in the following sequence:

(1) Flat plates tested cold.

(2) Welded plates tested at room temperature.
(3) Welded plates tested at appropriate high temperature.

This would be followed by tests of pipe at high temperature.

3.3.2. Penetrant Testing (PT), Magnetic Particle Inspection (MPI) and Alternating
Current Potential Drop (ACPD)
3.3.2.1. Description of equipment used
The dye penetrant used was a commercial product, Checkmor-C, made by Ely Chemicals, in
conjunction with developer LD-7. This is claimed to be suitable for testing at high
temperatures and was used in all cases to indicate crack initiation during the tests. Whilst, in
the field, dye penetrant has limitations, in the controlled conditions of the laboratory it would
be a reasonable indicator of cracking.

MPI was only applied to the DSS specimens at room temperature after completion of the
SCC tests. This was the main preliminary characterisation method for these samples. A
Magnaflux Y6 yoke was used to provide the magnetic field, and this was used in conjunction
with Magnaflux Type 7HF black ink.

ACPD was used to obtain an estimate of crack depth, and was used on all samples at room
temperature. The equipment used was a Karl Deutsch Type RMG 405.

3.3.2.2. Procedure (plates and pipes)

The dye penetrant was used as per the manufacturer’s instructions, during the crack initiation
stages for intermittent inspection whilst retaining the specimen in the loading jig. Preliminary
cleaning of the surface was with water (to remove the salt deposits). The soak time for the
penetrant was one hour.

The MPI was carried out on the DSS plates after they had been removed from the rig. Since
the crack location and primary orientation was known, the yoke was set at right angles to this
direction. The result was photographed, and the cracks then identified for subsequent
measurement by ACPD.

The ACPD system was set up according to the manufacturer’s instructions on a mild steel test
block.

15558/6/06
Copyright © 2006, TWI Ltd
3.3.2.3. 316L stainless steel welded plates
These were tested by dye penetrant only. The indications were too small for ACPD.

3.3.3. Acoustic Emission (AE)

Acoustic emission works by detecting the noise generated by incremental growth in cracks.
AE was applied for the hot situation only. This work was supervised and equipment provided
by Lloyd’s Register, primarily for the full pipe tests. However, as a setting up and training
exercise, monitoring using acoustic emission was carried out on the 316L and 22Cr duplex
stainless steel plates. Four sensors were mounted on the top surface of cold unstressed plates
away from the highly stressed area (Fig.7). Since the sensors were rated for use up to 140°C
(compared to a 150°C test temperature) the heating system had to be modified as described in
section 3.2.1.

Tests were carried out on Plates 316L-01, 2205-03, 2205-04 and 2205-05 and 316LW1-1
weld as well as the pipe samples. Data were continuously recorded and sent to Lloyd’s
Register periodically for analysis.

The tests on the pipes were more comprehensive with additional sensors placed to cover a
wider frequency range (100KHz to 800KHz). A more complete description of the procedure
employed is given in Appendix C.

3.3.4. Comparative Vacuum Monitoring (CVM)

Comparative vacuum monitoring (CVM) works by measuring the change of the vacuum
pressure inside plastic sensors. Surface breaking cracks that cross the vacuum sensor(s) cause
a change to the vacuum inside the sensors. The CVM system was designed to give an audible
warning signal when a crack was detected. A threshold pressure of 300Pa was set for this
purpose.

Figure 8a shows a cross section of a sensor. Figure 8b shows the principle of the method.
These sensors were applied on the top surface of cold plates prior to mechanical loading.
Sealing of the sensors was aided by silicone and grinding and cleaning of the plate surface
was carried out to remove discontinuities from plate manufacture and any contamination.
Grinding was performed to a 600 grit finish. The area inspected was 160mm long and was
located at both sides of the longitudinal centre line as show in Fig.8c. Figure 9 shows how
this appears in practice.

Three CVM sensors were applied on each plate and the changes in pressure were
continuously recorded and the relevant data were stored electronically. The CVM sensors
were applied on both 316L and 22Cr duplex stainless steel plates. The drop evaporation test
was stopped each time an audible signal was created and the plate surface was examined
using dye penetrant to verify detection of any SCC.

3.3.5. Eddy Current Testing

3.3.5.1. Scope
All plates and pipes were tested by the eddy current method. It was expected that, in order to
operate at high temperatures, the probe would need to be operated at a distance from the

15558/6/06
Copyright © 2006, TWI Ltd
surface. The experiments were therefore separated into two components: (1) performance of
different probes with lift-off, and (2) heat insulating performance of materials that could be
inserted between the surface and the probe.

3.3.5.2. Description of equipment used

A Hocking Phasec 2200 instrument was used for these experiments. The following probes
were used:

- 102P1, 2MHz, NFe absolute shielded probe.

- 801P15L, 100kHz, Weldscan probe.
- 801P02, 100kHz, Weldscan probe.
- 800P01JD1P, 100kHz, Weldscan probe.
- 800P01, 100kHz, Weldscan probe.

3.3.5.3. Procedure
The equipment was set up for the first probe with lift off signal horizontal, and in the other
cases with the slot signal vertical. The difference is due to the nature of the probe, the first
probe has a large lift off signal that needs to be eliminated from the analysis, the Weldscan
probes are differential and in theory have no lift off signal.
It was established that the duplex plate in use was similar electromagnetically to mild steel at
the frequencies used. The sensitivity was set such that a 1mm deep slot in a steel test block
(for the duplex) and a stainless steel one for the 316L plate was around a 40% screen height.
After the initial sensitivity check on a test block, the general procedure for scanning the plates
was to mark the plate in 5mm intervals and to scan along these paths. This was restricted to
known cracked areas for the 316L samples. The vertical component of the signal was then
recorded as the probe was scanned across a plate. This scan was repeated along the plate to
establish sensitivity to the small cracks and to observe the spatial resolution.

3.3.5.4. Variation of performance with lift off

The scanning operation was carried out as above and the lift off was varied by inserting
suitable spacers between the probe and the metal. All scans were recorded in a database. Data
was extracted to allow the analysis of probe performance with lift-off. This analysis consisted
of assessing the vertical axis amplitude of the eddy current signal obtained, with different
values of lift-off at the same positions on each plate.

3.3.5.5. Adaptation for high temperatures

In order to operate the probe in a normal scan, the use of a heat insulating pad between the
probe and the surface of the metal was investigated. The following different proprietary
insulating materials were used:

- Ceramic:Duratec, thickness from 0.5mm to 6mm.

- Superwool paper (flexible insulating), thickness: 1 and 2mm.
- Barrikade, thickness: about 2mm (TWI proprietary heat insulating material).

15558/6/06
Copyright © 2006, TWI Ltd
The time taken for the temperature at the probe to rise above 60 °C was noted. Figure 10
gives the results of this experiment.

It was noted that the times available for the inspection before the temperature became too
high were too short to be useable (less than 2 minutes). It was therefore decided to test
whether a simple air gap, or cool air flow, could keep the probe cooler for longer. The results
are shown in Fig.11. On the basis of these results, it was therefore decided to manufacture a
probe holder that contained an air supply. This is shown in Fig.12.

The adapted eddy current probe (Fig.13) was applied firstly to a plate sample that was heated.
It was then deployed to the welded DSS pipe while hot (Fig.14). Using the device on a hot
weld was quite difficult due to the lack of tactile information usually available to the operator
(especially to identify the weld toe position). Therefore although the principle of the device
has been proved, a better design should be derived for practical use in the field.

3.3.6. Alternating Current Field Measurement (ACFM)

3.3.6.1. Scope
The ACFM method was used to test only the duplex samples.

3.3.6.2. Description of equipment used

A TSC Amigo instrument was used, together with two different probe types (Type 271 Mini
Probe and a weld probe with encoder type 325).

3.3.6.3. Procedure
The TSC system automatically provides a probe file and sensitivity set-ups for ferritic steel.
These were found to work well on the duplex for detection.

The mini-probe was used for testing the plates both hot and cold. The weld probe with the
encoder was useful to demonstrate the use of an array but was not a serious candidate for the
hot tests because of the encoder.

The pipes were tested cold with both probes. A further test at a lift off of 2mm was also
carried out with the mini probe to check what would happen should a hot test be carried out.

3.3.6.4. Adaptation to hot testing

An adaptation to hot testing similar to that for the eddy current probe was carried out as
shown in Fig.15. This was deployed on a hot plate as shown in Fig.16.

3.3.7. Ultrasonic Testing

3.3.7.1. Scope
Ultrasonic Testing was carried out on all plate and pipe samples at room temperature.

15558/6/06
Copyright © 2006, TWI Ltd
3.3.7.2. Description of equipment used
An R/D Tech Focus instrument was used, with Probe Type Type Imasonic 7MHz 32 element
integral wedge. The sensitivity was set using a 3mm side drilled hole at an appropriate depth.

3.3.7.3. Procedure
The objective of using phased array ultrasonics was to check the detectability of the cracks if
they had appeared on the inside of the pipe. Therefore the ultrasonic beam was skipped on to
the surface. The phased array system was also used to fingerprint the welded samples prior to
cracking.

The signal was recorded by scanning along the plate or pipe with an encoder to indicate the
distance travelled. It was sometimes necessary to do this at a number of positions (parallel to
the cracks) on the duplex plate because the cracks were spread over a wide area.

The phased array ultrasonics using the same probe were also applied to the pipe samples.

3.3.8. Eddy Current Array

3.3.8.1. Scope
Two scans were made with eddy current arrays.

3.3.8.2. Description of equipment used

An R/D Tech MS5800 was employed, with two different array probes used for the different
materials. For the 316L stainless steel, a probe with 8 elements (2 rows of 4) was used. For
the duplex plate a probe with 16 elements (two rows of 8) was used. Both were operated at a
frequency of 1MHz. The smaller probe had a smaller element size.

3.3.8.3. Procedure
The system was set up as per the manufacturer’s instructions on appropriate test blocks.

3.3.9. Pulsed Eddy Current

The experiments on pulsed eddy current were carried out by the University of Huddersfield,
using their own system and two plates were loaned to them (Plates 2205-03 and 2205-05) for
the purpose. A full report of their work is given in Appendix G.

3.3.9.1. Description of equipment used

A GMR (Giant Magnetoresistive) probe was used for the experiments with a 1.5A excitation
current.

3.3.9.2. Procedure
The probe was scanned across the cracked area in three locations. It was applied to both the
cracked and uncracked side of the specimens. This is described in Appendix G.

15558/6/06
Copyright © 2006, TWI Ltd
3.3.10. EMAT and Laser Ultrasound Tests
3.3.10.1. Scope
This work was carried out by the University of Warwick. The same two plates were used for
these tests as for the pulsed eddy current tests.

3.3.10.2. Description of equipment used

Appendix H gives the main details of the experiments carried out. Two systems were used.
The EMAT system was a transmit-receive system. This equipment generated a Rayleigh
wave through the specimen with a frequency of 200KHz. This frequency is lower than is
normally used for UT. The plates were slightly small for the tests but useful results were
obtained. A laser unit was also used as an alternative signal generator.

3.3.10.3. Procedure
The system was scanned across the specimen in the same locations as the pulsed eddy current
described above.

3.3.11. Thermal Methods

Thermography works by inducing a localised source of heat and observing the temperature
effect as the heat transmits along the plate surface. Cracks in the surface are expected to
cause disruptions in the heat flow, and therefore non-linearities in the temperature gradient.

A hot air source was used to heat one end of a plate while the surface temperature was
monitored with a thermal camera. Thermal cameras are usually designed to work at large
distances; therefore they can be out of focus at small distances. In an attempt to overcome
this limitation, a Fluke Type Ti 30 camera with a close-up lens was employed.

A preliminary feasibility study on plate 2205-05, a heavily cracked plate, was carried out,
together with some work on a fatigue cracked sample that had been used in previous work
(for reference).

3.3.12. Laser Shearography

Two plates were examined by Dantec Ettemeyer for TWI, one was a duplex plate containing
large cracks (2205-02) the other was of 316L containing smaller cracks (316L-02). Laser
shearography requires that the sample be observed in two separate strain states. To achieve
this the samples were placed in the bend test rig. Appendix I gives the details of the method
and the results.

4. RESULTS
4.1. HOT TENSILE TESTS
The stress-strain curves from the hot tensile tests are shown in Fig.1. The average 0.2% proof
stress of UNS S31603 (316L) material at 70°C was 221MPa and the average 0.2% proof
stress of the UNS S32205 was 377MPa at 150ºC.

10

15558/6/06
Copyright © 2006, TWI Ltd
4.2. MECHANICAL LOADING AND DROP EVAPORATION TESTS
4.2.1. Austenitic Stainless Steel Plates
4.2.1.1. Loading and drop evaluation
Plain austenitic stainless steel plates 316L-01 and 316L-02 were loaded to maximum strains
of 4445µε and 5058µε, respectively. The welded plates 316L-W1 and 316LW1-1 were
loaded to 3680µε and 5248µε, respectively.

These strains were above 3200µε which corresponded to the 0.2% proof stress of the material
at 70°C and thus it was expected that the maximum stress on the top surface of the plates was
higher than the 0.2% proof stress of 316L material.

Salt deposits formed in the vicinity of the droplet splash area (Fig.17a). Figure 17b shows one
plate 316L-02 at room temperature after examination using PT prior to unloading from the
test rig. Prior to PT, the salt deposit was carefully dissolved using warm water to protect the
integrity of the CVM and AE sensors. CVM sensors were initially used on plates 316L01 and
316L02, while AE sensors were used for plate 316L01 only.

A summary of loading conditions is shown in Table 2.

4.2.1.2. Dye penetrant inspection

The plain austenitic stainless steel plates exhibited stress corrosion cracking. This was
detected on 316L-01 using PT, 20 days after start of exposure. A short crack, approximately
5mm long, was found on the surface of the plate under salt deposits. Numerous pits were
observed in the vicinity of the cracks (Fig.18).

Two SCC cracks were detected on the surface of plate 316L-02 using PT, 24 days after start
of exposure. The lengths of the cracks were 3.8mm and 7.8mm, respectively, and they were
observed under the salt deposits close to the centre of the plate (Fig.17b). No cracks were
detected in the centre of the plate under the drippers, where salt deposits did not form.

Penetrant testing of the welded austenitic plates 316LW1-1 and 316LW1-2 showed small
indications (Fig.19) in the parent material 62days and 14days, respectively, after start of
exposure. No cracks were detected in the weld metal.

4.2.2. Duplex Stainless Steel Plates

4.2.2.1. Loading and drop evaporation
Plates 2205-02, 2205-03, 2205-04, 2205-05 and 2205-06 were loaded to a maximum strain of
6031µε, 6100µε, 6000µε, 6004µε and 5500µε, respectively. These values of strain were
above 4000µε which corresponded to the 0.2% proof stress of the material at 150°C and it
was expected, therefore, that the maximum stress on the top surface of the plates was higher
than the 0.2% proof stress of the 2205 material.

Salt deposits formed immediately below the drop-splash area (Fig.20).

11

15558/6/06
Copyright © 2006, TWI Ltd
The fracture faces of SCC cracks in plate 2205-4, 316-01 and pipe 1 and 2 showed a brittle
morphology with negligible ductility (Fig.37-39).

4.4. NDE METHODS

4.4.1. General
A summary showing the minimum crack depth (measured by ACPD) which was detected
(where the signal was identifiable with the crack), by each technique, over the range of
samples is given in Table 4.

4.4.2. Dye Penetrant Inspection /MPI/ACPD

Both austenitic stainless steel plates and duplex stainless steel plates exhibited stress
corrosion cracking. This was detected on 316L-01 using PT, 20 days after start of exposure.
A short crack, approximately 5mm long, was found on the surface of the plate under salt
deposits. Numerous pits were observed in the vicinity of the cracks (Fig.18)

Two SCC cracks were detected on the surface of plate 316L-02 using PT, 24 days after start
of exposure. The lengths of the cracks were 3.8mm and 7.8mm, respectively, and they were
observed under the salty deposits close to the centre of the plate. No cracks were detected in
the centre of the plate under the drippers, where salty deposits did not form.

The DSS plates were examined with MPI and ACPD after initial detection visually and by
PT. A detailed set of results is given in Appendix B. Figure 40 shows a typical MPI result
with the corresponding ACPD data in Table 5. The cracks on the 316L stainless steel samples
were rather small and only single ACPD readings were possible.

Table 6 shows a comparison of crack sizes and ACPD readings. The crack size is given two
values. One is the maximum distance of a crack tip normal to the surface (called depth), the
other is the distance from the surface breaking point of the crack, along the crack, to the tip
(called the crack face length). This is the measurement made by ACPD. Fig.41 shows the
results from Table 6 diagrammatically, except for three points that had an ACPD reading
greater than the plate thickness.

4.4.3. Acoustic Emission

Table 7 shows the times when there was an increase in the rate of acoustic emission signals
(possible indication of cracking) compared with when cracking was detected by the dye
penetrant tests.

Appendix C gives a detailed report on the acoustic emission tests. Although many plates were
tested, the small size of the samples made detection of individual bursts of the AE
impossible.

On the pipe tests, better results were obtained and the acoustic emission signal exceeded the
background noise almost as soon as the tests began. This was more apparent in the second
test where the AE signal exceeded the noise by 20dB.

13

15558/6/06
Copyright © 2006, TWI Ltd
4.4.4. Comparative Vacuum Monitoring
No results were obtained from the CVM on the 316L plates, as the cracks didn’t pass under
the sensors. The CVM system was used on Plates 2205-02 (as a trial) and 2205-04. Some
difficulties were experienced because the salt formation seemed to get under the sensor,
producing a reading due to loss of vacuum and then block the sensor with a deposit of salt.
Figure 42 shows this effect.

By improving the sealant deposition on the second sample and with a wider footprint on the
sensor this problem was substantially eliminated. The system was used to trigger the end of
the test (Fig.43). However, cracking was extensive before the system triggered. This
presumably reflects the protection of the underlying material by the CVM sensors.

It may be that this system will be less subject to the environment in normal operational
circumstances than it is in these accelerated tests where large salt deposits form.

4.4.5. Eddy Current Inspection

The full eddy current inspection results are given in Appendix D. Figures 44 and 45 shows
the typical signal from such a scan. Figure 46 shows a comparison of the resolution of a weld
probe and a pencil probe with that from MPI when the probe is applied directly at the surface.
Figure 47 shows the loss of signal as lift off is increased. It can be seen that the weld probes
generally have a similar performance, the 800P01 probe was chosen for adaptation to the hot
surfaces because the larger probe 801P15L was a right angled tip and was unsuitable.
Appendix D gives the remaining scans of the plates and welded plates.
Table 8 (a and b) show a comparison between all the eddy current probes for signal size
against crack face length measured from the metallographic sections on. Fig 48 shows this
information graphically for Plate 2205-05.
Figure 49 shows examples of the scans taken on the hot pipe with the eddy current probe. The
crack can be distinguished but only by distortion of the weld toe indication.

4.4.6. Alternating Current Field Measurement (ACFM)

The ACFM system showed good detection of the cracks on the duplex samples. Figure 50
shows an example of the scan of a duplex plate with the mini probe, Figure 51 shows a scan
of the same area with the array probe. Figure 52 shows the signal from the pipe with no lift
and Figure 53 the same area with 2mm lift off. The adapted probe for high temperatures was
tested on a hot plate and the results are shown in Fig.54.

The full set of ACFM results is given in Appendix E.

4.4.7. Ultrasonic Testing (UT)

Figure 55 shows an example of detection of a small crack in plate 316L-01 (see Fig.18). The
crack indication is clearly seen, but this appears to be a corner echo with no indication of
height. Figure 56 shows an image from multiple cracking of duplex Plate 2205-02. The main
cracks in this sample (see Fig.40) can be identified and some indications below the surface
showing crack depth are observable. The UT was also used to examine the pipe sample (cold)
and the results are shown in Fig.57. Some depth indications are present.

14

15558/6/06
Copyright © 2006, TWI Ltd
The full set of results from the ultrasonic testing are in Appendix F.

The UT testing carried out here has shown that the cracks are detectable in the simple plate
geometry. Detection of SCC cracks in a weld has not been investigated as the cracks formed
away from the weld. The phased array probes and wedges currently available cannot be used
at high temperature. However, standard ultrasonic probes with this temperature capability
have been manufactured and these could be used for inspection, although the images
produced may be less clear than the focussed beam images produced in these experiments.

4.4.8. Eddy Current Array

The eddy current array system was able to detect surface cracks in the stainless steel plate
(Fig.58). The larger probe used on the duplex material (Fig.59) shows how this method
provides a convenient way of scanning large areas of plate material for this type of cracking.
It might be expected that use of air cooling may also enable the technique to be deployed at
higher temperatures if needed.

4.4.9. Pulsed Eddy Current

Figure 60 shows that detection of the cracking in the DSS plate, when present as a group, was
possible when the probe was deployed on the same side as the crack. It had been hoped that
the low frequency components of the pulsed eddy currents might respond to the cracks when
deployed from the opposite face. This was not the case. Further work using magnetisation to
increase the skin depth penetration may produce some further benefit. A full report on the
experiments carried out is given in Appendix G.

The pulsed eddy current system may be deployed on hot material, with air cooling, however
it does seem that this method has no particular advantages over the commercially available
conventional eddy current and ACFM systems.

4.4.10. EMAT and Laser Ultrasound Tests

The EMAT tests showed detection of the cracks in the samples provided from both sides of
the sample. This is because of the low frequency used. The samples tested were not able to
give an idea of the limits of the sensitivity of the technique (except that cracks of the order of
3.5mm depth were detected). The technique is potentially capable of being used at high
temperatures, as it is non-contact. However, it is currently part of a research project and is not
available commercially.

4.4.11. Thermal Methods

The results from the thermal tests were disappointing. A fatigue cracked sample from another
project was used to show the principle of the technique. The result obtained showed that the
crack can be seen to distort the temperature profile as heat travels across the specimen. This
did not occur with the cracked duplex specimen. It is assumed that the tortuous path of the
stress corrosion cracks do not influence heat flow sufficiently to enable successful detection.

15

15558/6/06
Copyright © 2006, TWI Ltd
4.4.12. Laser Shearography
The laser shearography showed the potential to detect cracks of the order of 3mm deep or
greater (Fig.61). It is a non-contact method so could be used at high temperatures, and does
cover a wide area in one test, however it does need at least two separate stress states to give a
signal. It is also known to be subject to problems when vibration is present (some tests at
TWI had to be abandoned for this reason). These problems would undoubtedly reduce its
practical application.

5. DISCUSSION
5.1. PRODUCTION OF STRESS CORROSION CRACKS
The transgranular crack path, brittle fracture face morphology and crack branching are
evidence that the cracks produced were typical of stress corrosion cracks. The cracks in the
samples tested in three-point bend rigs were spread around the transverse centreline of the
plates where the strain was at a maximum, as may be expected. Cracking in pipes (four-point
bend tests) also developed at locations with maximum strain (12 o’clock position). The time
to cracking varied and, in general, it was longer when lower strain was applied, as would be
anticipated. This was evident from data on the DSS plates and pipes while the data from tests
on the 316L plates were more scattered.

The test conditions on DSS plates and pipes at 150°C were more aggressive compared to
those on austenitic plates at 70°C in order that the materials would be exposed to appropriate
conditions to induce SCC. The 150°C exposure appears to be particularly aggressive as
shown by the short times to cracking and the large number of cracks. Rapid evaporation of
natural seawater on the surface of the DSS material promoted formation of 'dry' salts and it is
likely that these salts maintained a very low pH compared to that of natural seawater (6-8),
which resulted in accelerated pitting and cracking in the substrate DSS plate. This obviously
was not the case for austenitic materials where cracking took longer to develop at the
comparatively lower test temperature (70°C) did not cause rapid evaporation and excessive
salt accumulation and/or drying, even though this is above the industrially accepted limit of
55°C for these grades.

The SDSS weld metal in DSS pipes 1 and 2 did not exhibit cracks. This observation was
consistent with published data showing that superduplex material is immune to chloride
induced stress corrosion cracking up to 250°C, approximately (6). Plates 316LW1-1 and
316LW1-2 also did not exhibit cracks in the weld metal implying that the larger specimen
thickness at the welds led to the stress being lower in these regions compared to the parent
material.

In summary, drop evaporation testing of plates and girth-welded pipes loaded in three-point
bend and four-point bend rigs provided repeatable results within the bounds of the strain and
temperature control applied for a damage mechanism which is corrosion driven and therefore
subject to some stochastic influence. The cracks produced were typical of stress corrosion
and so were appropriate for the basis of establishing effective NDE methods.

16

15558/6/06
Copyright © 2006, TWI Ltd
5.2. EFFECTIVENESS OF NDE METHODS IN DETECTING SCC
The acoustic emission and comparative vacuum monitoring both exhibited some success for
monitoring while cracks were being produced. In the case of AE there were occasions when
the signal showing cracking was detected prior to visual or dye penetrant indications (Table
7). Both AE and CVM were able to operate at high temperatures and detected cracks. Both
experienced some difficulty with the experimental conditions (the AE because of the small
plate size, the CVM because the salt deposits lifted the sensors).

The AE inspection of the pipes was relatively successful, and crack initiation appeared to be
detected, although the detection level was different in each case. This was probably simply
due to the different responses of the two cracks. A detailed description of the AE monitoring
trials is given in Appendix C.

One of the advantages of AE over other methods of inspection is its ability to scan large areas
and then locate problems areas. The trials showed that given a reasonably large specimen, the
location of the acoustic source was relatively accurate, although it was subject to some
scatter. Location was facilitated by the relatively simple geometry of the test specimens. In
more complex geometries location would be more difficult due to the changes in the wave
path, differences in the wave velocity, wave mode conversion and variation in the signal
attenuation associated with different wave types. For some common geometries it is possible
to develop the location algorithm to enhance location. However, in more complicated
geometries it may be possible to develop computer modelling techniques to increase the
accuracy of the source location.

It is not possible to provide statistically significant estimates of the probability of detection

for acoustic emission testing (or any other methods) based on the small sample size in this
project. It should also be noted that acoustic emission amplitude is related to the increase in
crack surface area rather than flaw size and in any case probability of detection (POD) is
determined by many factors such as background noise etc. However qualitative comments on
the relative POD of acoustic emission using low recording threshold compared to high
recording thresholds (required in the case of background noise) are contained in section 5 of
Appendix C.

The CVM method requires the crack to grow under its sensors. Ideally, the likely location of
the crack needs to be known, and examining a whole structure by this method might be
impractical, although perhaps the welds could be covered. The surface preparation also needs
to be good and this may present problems in practical application for offshore facilities.

Table 4 shows a summary of the smallest crack detected by MPI and dye penetrant and
measured by ACPD. The detection of these cracks by eddy current methods is also compared.
It is more difficult to compare the detection by other methods because individual cracks could
not be identified. (some qualitative data is given below) Table 4 shows that dye penetrant,
MPI, ACPD and eddy current methods (applied to the surface with 0mm lift off) all had a
similar sensitivity, but the lower frequency eddy current methods were less sensitive.

Figure 18 shows the sensitivity of the dye penetrant method for a surface crack in the 316L
material. Sectioning of this defect (Fig.27) showed a maximum through-wall crack height of
around 0.5mm. It was shown that eddy current (Fig.44), eddy current array (Fig.58) and

17

15558/6/06
Copyright © 2006, TWI Ltd
phased array ultrasonics (Fig.55) could detect this flaw relatively easily when deployed to the
surface. The same result could be expected, for the far surface case, for phased array UT.
Therefore for this plate, at temperatures up to the normal probe operating temperatures (about
60°C) these methods can be used instead of dye penetrant. This was a limited test of
sensitivity and further work (i.e. more samples with a wider range of cracks) would be
needed to establish more clearly the difference between the sensitivity of the techniques.

The MPI method (at room temperature) probably gave the best indication of cracks at the
surface of the duplex stainless steel, but does not have the advantages of ease of use and data
acquisition and recording. ACPD at room temperature gave an indication of crack severity.
Table 6 and Fig.41 shows that except for the anomalous readings greater than the wall
thickness, the ACPD reading was lower than the crack face length. There was a vague trend
for ACPD to increase with crack face length but the data was very scattered. On the pipe
samples, after the load had been removed the crack depth measured by ACPD was found to
be around 2mm in both pipes (Appendix B). This was less than that shown by the sectioning
(Figs 32-35). A possible explanation for this is that the cracks had electrical contacts across
their faces due to closure: a single reading of 7mm was obtained by the ACPD equipment
before the stress was taken off.

Crack depth measurement by ACPD is subject to a number of possible interfering effects that
may affect the accuracy of the reading. The ACPD is only calibrated for a single crack, and
the feature measured is the crack face length through the thickness (not the through thickness
depth). This is again only true for cracks where the face length is short compared with the
surface length. Electrical contacts across the crack faces will reduce the electrical resistance
and therefore reduce the reading. Additional branches will add to the reading, and errors will
be introduced if the contacts cover more than one crack. Therefore complex cracking such as
branching SCC is unlikely to give accurate results, and it is not surprising that the ACPD and
section results are different.

In Table 8 and Fig.48 it can be seen that, although the cracks were detected, the amplitude of
the eddy current response was very scattered compared to the crack face length. The reasons
for this are similar to those mentioned above for ACPD.

For the duplex stainless steel material the high frequency eddy current method could
discriminate cracks at the surface (Fig.46). However this probe type could not be used with
any lift off from the surface that might have protected it from the elevated temperature. Using
the weld probe or the ACFM method meant that crack groups rather than individual cracks
could be detected (Fig.46, 50 and 51). These methods could be used with lift-off and they
could be adapted for hot use relatively easily, with some anticipated loss of sensitivity
(compare, for example, Fig.52 and 53). The use of cold air flow to cool the probe was
effective, although for further use the design needs to be improved for easier manual
deployment in a practical application.

The phased array ultrasonics technique was also effective at detecting the crack groups from
the far side and gave some approximation of crack depth (although multiple echoes between
the cracks might make this interpretation difficult). Unfortunately this method could not be
used at high temperatures. Simpler high temperature probes are becoming available and may

18

15558/6/06
Copyright © 2006, TWI Ltd
possibly detect the cracks. They are unlikely, however, to be as good for crack
discrimination, but assessment would be warranted.

The cracks in the 316L welded samples developed away from the weld and therefore the
capability of the ultrasonic equipment to detect cracks in the large grain regions of the weld
was not assessed.

Of the remaining methods, the eddy current array (Fig.58 and 59) was suitable for detection
of the cracks when it was possible to operate with zero lift off. Optimising the parameters
may enable some lift-off and therefore allowing operation at elevated temperature to be
allowed in future developments.

The pulsed eddy current method was able to detect the surface cracks (Fig.60), but was not
able to detect cracks from the far side as had been hoped (i.e. for the detection of internal
cracks).

The EMAT system was effective at detecting the cracks from both sides of a plate (i.e.
allowing for the detection of internal and external cracking) with Rayleigh Waves (Appendix
H) and could potentially be used at high temperature. However, the sensitivity could not be
easily established from the samples tested.

The laser shearography seemed to be able to detect cracks of the order of 3mm deep (Fig.61)
but may be impractical in the field, because of the need to induce stress and to avoid
vibration.

The tight cracks in the duplex sample did not give a noticeable change in the heat flow
through the sample, as had been observed with a fatigue crack. This method does not appear
to be suitable as deployed.

6. CONCLUSIONS
1. Drop evaporation testing of austenitic type 316L and 22Cr duplex stainless steel
plates loaded in three-point bend, and girth-welded pipes loaded in four-point bend
test rigs provided a robust technique for studying SCC. For both 22Cr duplex and
316L materials, stress corrosion cracks always formed in the parent material or
HAZ outside of the welds.

2. The acoustic emission method was able to operate at high temperatures and
successfully detected cracking in the duplex stainless steel pipes. Tests involving
the plate samples were inconclusive, due in part to the small size of the plates.
Monitoring using acoustic emission to detect SCC can either be done by setting low
thresholds at the expense of increasing the number of spurious calls, or by setting
higher thresholds to eliminate spurious calls at the expense of possibly missing
some instances of SCC propagation.

3. The hot dye penetrant was able to detect cracks in both sample types to a high
sensitivity and at temperature. From the limited evidence available, for inspecting
316L stainless steel at temperatures up to 60°C, it could be replaced by eddy
currents (single probe or arrays), or phased array ultrasonics.

19

15558/6/06
Copyright © 2006, TWI Ltd
4. At higher temperatures (up to 150°C) it was found practical to cool the eddy current
weldscan probe and ACFM probes by an airflow. This however decreased the
sensitivity relative to hot dye penetrants. The phased array ultrasonics could not be
deployed at these temperatures.

5. If the electromagnetic or ultrasonic methods are to replace dye penetrant for

inspection of austenitic type 316L or 22Cr duplex stainless steels at low
temperatures then comparison over a wider range of crack sizes is needed. For
application to hot surfaces, the ACFM and eddy current methods used in this project
would need further development. Pulsed eddy current testing offered no advantage
over conventional methods.

6. The EMAT system could detect cracks in plate samples from both the top and
bottom faces by Rayleigh waves, and could possibly be developed for hot
application. The Rayleigh wave method might also offer a method of detecting sub-
surface flaws, due to internal cracking.

7. The MPI method could be used on the DSS specimens for delineating the cracks at
room temperature. This was not possible at high temperatures. ACPD was used
successfully to give an indication of crack severity, but undersized the cracks in the
duplex stainless steel pipe.

8. The comparative vaccum, shearography and thermography methods are not suitable
for the detection of SCC.

7. RECOMMENDATIONS
For the detection of the propagation of stress corrosion cracks, acoustic emission should be
considered. However, further refinement may be necessary to determine safe alarm
thresholds that allow any significant crack growth to be detected, whilst not triggering an
undue number of false calls.

For stress corrosion cracking at moderate temperatures in austenitic stainless steels, eddy
currents (single probe or arrays), or phased array ultrasonics should be considered as
complementary to hot penetrant testing. However, comparison over a wider range of crack
sizes is needed to better establish real time performance.

At temperatures significantly greater than 60°C (150°C in the present work), hot penetrant
testing proved to offer superior sensitivity to electromagnetic (ACFM or eddy current)
methods using cooled probes. Improved cooling designs could reduce the difference in
sensitivity and should be investigated where use of dye penetrant is inconvenient.

The Rayleigh wave method should be further investigated as a means of detecting sub-
surface flaws.

Further work needs to be carried out on a wider range of realistic samples if a statistical level
of confidence of the capability of any technique is required.

20

15558/6/06
Copyright © 2006, TWI Ltd
8. ACKNOWLEDGEMENTS
The authors acknowledge the contributions of Lee Smith, Stuart Bond, Mike Bennett, Kim
Hayward and Katy White.

9. REFERENCES
1. 'Surface inspection methods for detection of stress corrosion cracking in austenitic
and duplex stainless steels. First progress report', TWI Report 15558/01/05.

2. Notes of First Sponsor Group Meeting held on 17 May 2005 at TWI Ltd, Granta
Park, Great Abington, Cambridge, UK.

3. Bond S, Leonard A and Rudlin J, 'Surface inspection methods for detection of stress
corrosion cracking in austenitic and duplex stainless steel', TWI Proposal PR6899-1,
August 2003.

4. Minutes of launch meeting, TWI, Granta Park, Great Abington, Cambridge, July
2003.

5. Minutes of inaugural meeting, TWI, Granta Park, Great Abington, Cambridge,

September 2003.

6. Gunn R: ' Duplex stainless steels, Microstructure, properties and applications',

Woodhead Publishing Ltd. Abington Hall, Abington, Cambridge, 1997.

7. Steinsmo U and Drugli J M: 'Assessment of susceptibility to chloride stress

corrosion cracking of highly alloyed stainless steels. Patr1: Drop evaporation test
method', Corrosion 97, paper 193.

8. Drugli J M and Steinsmo U: 'Assessment of susceptibility to chloride stress

corrosion cracking of highly alloyed stainless steels. Part 2: A new immersion test
method', Corrosion 97, Paper 194.

21

15558/6/06
Copyright © 2006, TWI Ltd
Table 1 Chemical analysis data and compositional specification of plate and pipe materials (TWI analysis references S/05/17, O/N 05/3,
S/02/307 and O/N02/35).

Element (wt%)
Material C Si Mn P S Cr Mo Ni Al As B
316L Plate 0.017 0.32 1.49 0.029 0.007 17.6 1.96 11.0 <0.01 0.012 <0.001
22Cr duplex plate 0.015 0.43 0.81 0.024 <0.002 22.6 3.10 5.9 0.01 0.010 ….
22Cr duplex pipes 0.022 0.5 1.44 0.019 0.002 22.4 3.26 5.7 - -
UNS S31603 austenitic 0.030 0.75 2.00 0.045 0.030 16.0- 2.00- 10.0- …. …. ….
stainless steel* max max max max max 18.0 3.00 14.0
UNS S32205 duplex 0.030 1.00 2.00 0.030 0.020 22.0- 3.0-3.5 4.5- …. …. ….
stainless steel* max max max max max 23.0 6.5

Element (wt%)
Sample ID Nb Pb Sn Ti V W O N Co Cu
316L Plate <0.01 <0.002 0.006 <0.005 0.03 <0.05 0.009 0.042 0.17 0.20
22Cr duplex plate <0.01 <0.002 0.007 <0.005 0.13 <0.05 0.002 0.16 0.11 0.20
22Cr duplex pipes - - - 0.001 0.11 <0.05 0.004 0.17 - 0.01
UNS S31603 0.10
austenitic stainless …. …. …. …. …. …. …. max …. ….
steel*
UNS 32205 duplex
stainless steel* …. …. …. …. …. …. …. 0.14- …. ….
0.20

* Composition according to ASTM A240/240M-03c.

…. Not specified.
- Not determined

22

15558/6/06
Copyright © 2006, TWI Ltd
Table 2 Summary of test conditions employed during the three-point and four-point bend tests of plates and girth welded pipes, together with %
ferrite and ASTM grain size.

Specimen Type Temperature Maximum Type of test Time lapsed %Ferrite ASTM Time lapsed between drip
°C Strain between ± 95%CI* Grain start and detection of crack
(µε) mechanical Size with PT
loading and drip
start
Duplex Plate 2 150 6031 - 46±7 11 -
DuplexPlate 3 150 6100 - 46±7 11 13 days
Duplex Plate 4 150 6000 - 46±7 11 14 days
Duplex Plate 5 150 6004 Three-point - 46±7 11 4 days
Duplex Plate 6 150 5500 bend - 46±7 11 2 days
316L Plate 1 70 4445 - - 6 20 days
316L Plate 2 70 5058 - - 6 24 days
316L Weld 1 70 3680 - - 6 62 days
316L Weld 2 85 5248 - - 6 14 days
Duplex Pipe 1 150 6080 Four-point bend 50h 53±7 9 1 day
Duplex Pipe 2 150 5160 50h 53±7 9 4.8 days

* 95%CI (Confidence interval) = 2σ/√n-1 where σ is the standard deviation and n the number of fields measured.

23

15558/6/06
Copyright © 2006, TWI Ltd
Table 3 Summary of NDE methods.

PAUT EC EC Array ACFM PEC EMAT/

Fingerprint DP/MPI ACPD Shearography Thermography
Note1 Man Note1 Note1 Note1 Laser UT
Duplex Plate 2 9 9 9 9 9 9
DuplexPlate 3 9 9 9 9 9 9 9
Duplex Plate 4 9 9 9 9 9
Duplex Plate 5 9 9 9 9 9 9 (Hot) 9 9 9
Duplex Plate 6 9 9 9 9
316L Plate 1 9 9 9 9 9
316L Plate 2 9 9 9 9 9
316L Weld 1 9 9 9 9
316L Weld 2 9 9 9 9
Duplex Pipe 1 9 9 (Hot) 9 9 9 9
(Hot)
Duplex Pipe 2 9 9 (Hot) 9 9 9 9

Note1: Unable to identify individual small cracks

24

15558/6/06
Copyright © 2006, TWI Ltd
Table 4 NDE detection of known cracks.

DP (SS) / MPI (duplex) ACPD Readings

EC Man
Crack ID Scan Num. (mm)
Duplex Plate 2 2 13 0.1 detected
DuplexPlate 3 6 13.5 0.2 detected
Duplex Plate 4 7 16 0.3 detected
Duplex Plate 5 8 14 0.7 detected
Duplex Plate 6 18 9 0.1 detected
316 Plate 1 1 1 0.3 detected
316 Plate 2 0.8 1 defect detected
316 Weld 1 2 defects detected
316 Weld 2 3 defects detected
Duplex Pipe 1 1 2 4.2 detected
Duplex Pipe 2 2 1 1.3 detected

NB: Eddy current results with high frequency probe

25

15558/6/06
Copyright © 2006, TWI Ltd
Table 5 ACPD readings from cracks found with MPI on duplex plate 2205-03.

Crack ID Length (mm) Scan number ACPD readings (mm)

6 1.1
1 12 6.5 2.1
7 2.6
6 1
2 10 6.5 2.4
7 3.6
8 4.6
9 5
3 24
10 3.8
11 1.3
13 1.4
5 10
13.5 1.8
6 4 13.5 0.2
14 1.3
9 12
15 0.8
10 8 14.5 0.4
17.5 1.8
12 8
18 1.4
19 4 19 0.5
22 2.4
23 6.5
21 27 24 6.9
25 3
26 1.3

26

15558/6/06
Copyright © 2006, TWI Ltd
 Table 6 ACPD and metallographic measurement on sectioned samples.

Metallographic readings
Specimen Metallographic Scan Cracks
ACPD readings Depth* Crack face
ID section No ID
length*
2205-04 met3 21 9 4.8 3 6.4
Additional section 3 18.7 8 16
10
(met6) 8 7.4 5 12 (7.5)
Additional section 10 3.4 5.5 12
2205-05 (met 7) 15 3 37.7 9 20
9 28.7 6 14
Additional section 3 6.2 8.5 Variable
25
(met 8) 11 1.1 3 7
met4 1 1 9.5 7 9.3
DSS Pipe1 Additional section 7 9
2 1 4.2
(met 9)
met5 1 1 4.7 5.5 ≥5.5
1 2 1.3 5 6
DSS Pipe2
Additional section 6 (6)
2 1 2.2
(met10)
Additional section 0.3
316L-01 1 1 0.3-0.5
(met11)

All readings are given in millimetre

Depth*: Maximum distance of a crack tip normal to the surface.
Crack face length*: Distance from the surface breaking point of the crack along the
crack, to the tip.

27

15558/6/06
Copyright © 2006, TWI Ltd
 Table 7 Time of crack detection by AE and DP.

Sample Time Detected AE Time detected DP

Duplex plate 4 29/3 4/4
Duplex plate 5 0200 on 21/8 22/8 extensive cracking
Duplex Pipe 1 0100 on 05/12 13.20 on 6/12 (ACPD =7mm)
Duplex Pipe 2 21.00 on 09/01 Not detected by DP 10/1. Detected 11/1 morning

Table 8 Eddy current results on sectioned samples.

(a): Results obtained on the plate 2205-05.
P102P1, 2MHz P122P1D, 2MHz 123P1D, 2MHz 800P01, 100kHz 801P15L, 100kHz 800P04, 100kHz 800P04, 300 kHz 800P04, 600 kHz
Crack face

231°, 13dB 180°, 17dB 189°, 22dB 134°, 40dB 295°, 30dB 280°, 28dB 168°, 31dB 70°, 37dB
Crack ID
Scan No

Length
(mm)

Lift off (mm)

0 0.5 1 0 0.5 1 0 0.5 1 0 0.5 1 0 0.5 1 0 0.5 1 0 0.5 1 0 0.5 1

10 3 16 70 25 <10 max 50 35 max 65 35 45 40 25 45 45 25 45 40 25 65 45 40 65 45 40

8 12 50 15 <10 80 35 25 80 45 30 40 25 20 40 30 <10 30 20 20 45 40 30 45 40 30
15 10 12 45 25 65 25 65 55
3 20 55 25 <10* 90 35 25* 90 55 35* 45* 40* 25* 45* 40* 30* 45* 40* 25* 50* 45* 40* 50* 45* 45*
9 14 50 80 35 80 55
25 3 Variable 45 <10 <10 70 25 15 70 35 25 45 25 20 40 25 20 40 25 20 45 40 25 45 40 30
11 7 10 <10 <10 20 <10 <10 20 <10 <10 <10 <10 <10 <10 <10 <10
(b): Results obtained on the duplex pipes and plate 316-01
Sample ID Scan No Crack ID Crack face P102P1, 2MHz Settings
length (mm)
Pipe1 1 1 9.3 Not done The amplitude of the Eddy current signal is given in % screen height
210°, 20dB
2 1 9 60 Calibration is for 1mm=40% screen height at 0mm lift off
Pipe2 1 1 ≥5.5 Not detected All measurement have been made with the same gain for each lift off
1 2 6 75 210°, 20dB with one probe
2 1 6 55 *: Cracks can not be identified from the other
316L-01 1 1 40 212°, 20dB

28

15558/6/06
Copyright © 2006, TWI Ltd
 250

200

150
Stress, MPa

100

50

0
0 1000 2000 3000 4000 5000 6000 7000
Strain, microstrain

a)

450

400

350

300
Stress MPa

250

200

150

100

50

0
0 1000 2000 3000 4000 5000 6000 7000 8000
Strain, microstrain

b)

Fig.1 Stress-strain curves from hot tensile tests in air. Dotted lines highlight the 0.2%
proof stress.

a) 316L type austenitic stainless steel tested at 70°C.

b) 22% Cr duplex stainless steel tested at 150°C.

29

15558/6/06
Copyright © 2006, TWI Ltd
 CVM Sensors
Strain gauge

150mm
Seawater
Drippers

a)
Acoustic emission sensors

60mm

300mm
100
32mm
12mm

HEATER HEATER

200mm

32mm

32mm

455mm
b)

Fig.2 Sketch showing the three-point bend rig and attached sensors.

a) Plan view.
b) Side view.

30

15558/6/06
Copyright © 2006, TWI Ltd
 Seawater Drippers CVM sensors

40mm

a) 15888_316_2_b

Acoustic emission sensor

Electric heaters
Hot air nozzles

40mm

b) 15888_2205_4_a

Fig.3 Photographs showing as-loaded test plates in three-point bend rigs with CVM
and acoustic emission sensors attached.

a) 316L type stainless steel plate (no AE).

b) 22Cr duplex stainless steel plate.

31

15558/6/06
Copyright © 2006, TWI Ltd
c) W1_1_20_9_205_0510

Fig.3 Cont’d Photographs showing as-loaded test plates in three-point bend rigs with
CVM and acoustic emission sensors attached.

c) 316LW1-1 welded austenitic stainless steel plate.

32

15558/6/06
Copyright © 2006, TWI Ltd
 seawater drips

169mm
thermocouples dia. OD
weld
12

9 3

100 mm
heating mantle 6
560 mm
a)

weld

15mm
12 o’clock
position

strain gauges

b)

weld

pipe

60 cm

300cm
c)

Fig.4 Sketch showing the four-point bend test of the girth welded duplex pipe.

a) Top view of pipe with heating mantles and thermocouples.

b) Detail of the top of the pipe (12 o’clock position) showing strain gauges.
c) Four point bend test arrangement.

33

15558/6/06
Copyright © 2006, TWI Ltd
 AE sensor

0.5 m

a) 6_12_2005_0063

1.5 cm

b) mcs_0006

Fig.5 Photographs showing as-loaded test pipe 1 in four-point bend rig with acoustic
emission sensors attached.

a) Overview
b) Greater magnification of the top of the pipe showing weld, strain gauges and
thermocouples.

34

15558/6/06
Copyright © 2006, TWI Ltd
 AE sensors

1m

a) 09_01_2006_0123

1.5 cm

b) _09_01_2006_0124

Fig.6 Photographs showing as-loaded test pipe 2 in four-point bend rig with acoustic
emission sensors attached. Scale shown for mid-point of test-piece.

a) Overview.
b) Greater magnification of the top of the pipe showing weld and strain gauges.

35

15558/6/06
Copyright © 2006, TWI Ltd
 2.5 cm

24_02_05_0138
Fig.7 Set up of AE sensors on plate.

36

15558/6/06
Copyright © 2006, TWI Ltd
 Channels for air flow

Adhesive bond to
surface

Fig.8(a) Cross section of CVM sensor.

Monitor Pump
(Pressure gauge)

Crack growing under

sensor
Sensor blocked at end

Fig.8(b) Principle of operation (pressure is low until crack causes leak path to
atmosphere).

C/L Strain gauges

Seawater
drippers
Sensors

150
7 C/L

80 80
Fig.8(c) Schematic showing the installation pattern of the CVM sensors on test plates. The
dimensions are shown in millimetres.

37

15558/6/06
Copyright © 2006, TWI Ltd
 2.5 cm

15558_316_2_6_4_2005_0266

Fig.9 Comparative vacuum monitoring sensors mounted on sample.

1.6
1.5
1.4
1.3
1.2
1.1
1 ‹ Ceramic
0.9 „ Ceramic + 1mm of superwool
Time (min)

0.8 ▲ Ceramic + 2mm of superwool

0.7 U Ceramic + 2mm of barrikade
0.6
* Ceramic + 2mm of barrikade +
0.5
1mm of superwhool
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5
Lift off (mm)

Fig.10 Time for different insulating materials and thicknesses (shown as lift-off) to reach
60oC at contact point.

38

15558/6/06
Copyright © 2006, TWI Ltd
 75
With cold airstream Without cold air
70

65

60

55

50
Temperature (C)

45

40

35

30

25

20

15

10

0
0 20 40 60 80 100 120 140 160 180 200
Time (s)

Fig.11 Effect of cooling airstream on Eddy current probe, temperature taken at two-
millimetre lift off from the surface plate (150ºC).

Eddy Current
Probe and
thermocouple

Air

Probe fixation

15559_probe adaptationdrawing_3D-3

Fig.12 Probe holder to contain airstream for eddy current probe.

39

15558/6/06
Copyright © 2006, TWI Ltd
 Air Supply

Thermocouple Probe

9.5 mm

15559_D006105_04

Fig.13 Air cooled Eddy current probe.

9.5 mm

15559_D006105_03
Fig.14 Probe deployed on hot weld.

40

15558/6/06
Copyright © 2006, TWI Ltd
 Air supply

Thermocouples

Sensor

1 cm

15559_D006127_05
Fig.15 ACFM Mini Probe adapted for hot use.

1 cm

15559_D006127_03
Fig.16 ACFM probe deployed on hot surface.
41

15558/6/06
Copyright © 2006, TWI Ltd
 10mm
a) 15558_316_2_6_4_05_0256

10mm
b) 15558_316_2_15_4_2005_0293

Fig.17 Photographs showing SCC cracks on the surface of austenitic stainless steel plate
316L-02.

a) Salt deposit prior to PT. Note salt discoloration.

b) SCC on the surface of the plate after PT. Note that cracks have not crossed the CVM
sensors.

42

15558/6/06
Copyright © 2006, TWI Ltd
 met1

met11

5mm

a) 15888_316_01_0165

5mm

b) 15888_316_01_0168

Fig.18 Photographs showing SCC cracks on the surface of austenitic stainless steel plate
316L-01.

a) PT-overview.
b) As above, but at greater magnification.

43

15558/6/06
Copyright © 2006, TWI Ltd
 met2

a) W1_0024

b) W1_2 15_12_2005_0099

Fig.19 Photographs showing the surface of welded austenitic stainless steel plates
following PT.

a) 316L-W1-1.
b) 316L-W1-2.

44

15558/6/06
Copyright © 2006, TWI Ltd
 35mm

a) 15558_2205_04_h

b) 15558_2205_3_20_4_2005_0310

Fig.20 Photograph showing accumulation of salt deposits on the surface of the 22Cr
duplex stainless steel plates prior to MPI. Note salt deposit accumulation in the drop splash
zone. Magnifications are given by marker and millimetre scale.

a) Plate 2205-04.
b) Plate 2205-05.

45

15558/6/06
Copyright © 2006, TWI Ltd
 5cm

15558_2205_02_Cracked duplex_0146

Fig.21 Photographs showing visible indication of SCC cracks on the surface of duplex
stainless steel plate 2205-04.

46

15558/6/06
Copyright © 2006, TWI Ltd
a) 15558_2205_3_20_4_2005_0318

b) 15558_2205_3_20_4_2005_0315

Fig.22 Photographs showing SCC cracks on the surface of duplex stainless steel plate
2205-03 after PT. Magnification is shown by half-millimetre scale.

a) Overview.
b) Crack propagating under the CVM sensor.

47

15558/6/06
Copyright © 2006, TWI Ltd
 TOP
A Plate 2205-4
Crack 1

Scan 5
Crack 3
Crack 2

Crack 4 Scan 10

Crack 5

Crack 6

Crack 7

Scan 15
Crack 8

Crack 9
Crack 8

Scan 20
met3

Crack 10

Scan 25

Crack 11 Crack 12

BOTTOM B
Fig.23 Photograph showing SCC cracks on the surface of duplex stainless steel plate 2205-
04 after MPI. Magnification is shown by millimetre scale. The locations where
metallographic sections were prepared are also shown, as are NDE scan positions

48

15558/6/06
Copyright © 2006, TWI Ltd
 TOP Plate 2205-5
A
Crack 1 Crack 2

Scan 5
Crack 3
Crack 4
Crack 5

Crack 6 Crack 7
Crack 8

met6
Scan 10

Crack 10 Crack 9

met7
Scan 15
Crack 11

Crack 12
Crack 13

Crack 14 Scan 20

met8 Scan 25

BOTTOM B
Fig.24 Photograph showing SCC cracks on the surface of duplex stainless steel plate 2205-
05 after MPI. Magnification is shown by millimetre scale. The locations where
metallographic sections were prepared are also shown, as are NDE scan positions.

49

15558/6/06
Copyright © 2006, TWI Ltd
 1 cm

a) 6_12_2005_0065

1 cm

b) 10_1_2006_0127

Fig.25 Photographs showing accumulation of salt deposits on the surface of the 22Cr
duplex stainless steel pipes prior to DPI. Note salt deposit accumulation in the drop splash
zone. Magnifications are given by marker and scale marks.

a) Pipe 1.
b) Pipe 2.

50

15558/6/06
Copyright © 2006, TWI Ltd
 met 9 met 4 4mm

1 1

a) NDE1

met 5 met 10

2 1

9mm

b) NDE2

Fig.26 Photographs showing SCC cracks on the surface of duplex stainless steel pipes.

a) MPI-overview on Pipe 1. Note numerous SCC cracks away from the weld.
b) MPI-overview on Pipe 2 Note numerous SCC cracks away from the weld.

51

15558/6/06
Copyright © 2006, TWI Ltd
 1

a) 2006-3-1-9-58-18-003

b) 2006-3-1-10-22-11-003

Fig.27 Photomicrographs showing stress corrosion cracking in sections met1 and met11
prepared from plate 316L-1.

a) Cracks in section met1- overview.

b) Greater magnification from region 1. Note multiple crack branches and transgranular
crack path.

52

15558/6/06
Copyright © 2006, TWI Ltd
c) 2006-6-8-11-13-57-003

Fig.27 Cont’d Photomicrographs showing stress corrosion cracking in section met1 and
met11 prepared from plate 316L-1.

c) Transgranular crack in section met11.

53

15558/6/06
Copyright © 2006, TWI Ltd
 1

3
a) 2006-2-21-10-53

b) 2006-2-21-10-42-17-003

Fig.28 Photomicrographs showing stress corrosion cracking in section met3 prepared from
plate 2205-04.

a) Overview.
b) Greater magnification from area 1 in Fig.28a. Note only small degree of crack
branching and transgranular crack path.

54

15558/6/06
Copyright © 2006, TWI Ltd
c) 2006-2-21-10-43-17-003

d) 2006-2-21-10-44-26-003

Fig.28 Cont’d Photomicrographs showing stress corrosion cracking in section met3

prepared from plate 2205-04.

c) Greater magnification from area 2 in Fig.28a. Note small fine crack branches and
transgranular crack path.
d) Greater magnification from crack tip (area 3 in Fig.28a).

55

15558/6/06
Copyright © 2006, TWI Ltd
 Crack 3 Crack 8

a) 2006-6-7-14-28-47-002

b) 006-6-7-14-27-10-003

Fig.29 Photomicrographs showing stress corrosion cracking in section met6 (scan 10)
prepared along from plate 2205-05.

a) Overview. Millimetre scale is shown.

b) Larger magnification of the area highlighted in Fig.29a. Note transgranular crack path.

56

15558/6/06
Copyright © 2006, TWI Ltd
 crack 10 crack 3 crack 9

2006-6-7-15-13-37-002

Fig.30 Photomicrographs showing stress corrosion cracking in section met7 (scan 15)
prepared from plate 2205-05. Millimetre scale is shown.

Crack 3

2006-6-7-14-8-51-002

Fig.31 Photomicrographs showing stress corrosion cracking in section met8 (scan 25)
prepared from plate 2205-05. Millimetre scale is shown.

57

15558/6/06
Copyright © 2006, TWI Ltd
 2 1

a) 2006-4-6-14-10-12-002

b) 2006-4-6-10-32-50-003

Fig.32 Photomicrographs showing stress corrosion cracking in section met4 prepared from
Pipe 1.

a) Overview.
b) Greater magnification from area 1 in Fig.32a. Note transgranular crack path.

58

15558/6/06
Copyright © 2006, TWI Ltd
 3

c) 2006-4-6-10-25-11-003

d) 2006-4-6-10-27-10-003

Fig.32 Cont’d Photomicrographs showing stress corrosion cracking in section met4

prepared from Pipe 1.

c) Greater magnification from area 2 in Fig.32a showing crack in the HAZ.

d) Greater magnification from area 3 in Fig.32c. Note transgranular crack path.

59

15558/6/06
Copyright © 2006, TWI Ltd
 crack 1

2006-6-7-12-9-36-002

Fig.33 Photomicrographs showing stress corrosion cracking in section met9 prepared from
Pipe 1. Millimetre scale is shown.

crack 1

2006-6-7-13-44-19-002

Fig.34 Photomicrographs showing stress corrosion cracking in section met10 prepared

from Pipe 2. Millimetre scale is shown.
60

15558/6/06
Copyright © 2006, TWI Ltd
 1

weld

a) 2006-4-6-14-13-13-002

b) 2006-4-3-15-11-23-003

Fig.35 Photomicrographs showing stress corrosion cracking in section met5 prepared from
Pipe 2.

a) Overview.
b) Greater magnification from area 1 in Fig.35a. Note transgranular crack path.

61

15558/6/06
Copyright © 2006, TWI Ltd
 HAZ

weld

c) 2006-4-3-15-15-17-003

d) 2006-4-3-15-17-4-003

Fig.35 Cont’d Photomicrographs showing stress corrosion cracking in section met5

prepared from Pipe 2.

c) Greater magnification from area 2 in Fig.35a, showing crack in the HAZ.

d) Greater magnification from region 3 in Fig.35c. Note crack branching and
transgranular crack path.

62

15558/6/06
Copyright © 2006, TWI Ltd
 1

a) 2006-4-6-14-47-13-002

b) 2006-4-6-14-29-48-003

Fig.36 Photomicrographs showing microstructure in section met2 prepared from plate

316LW1-1.

a) Overview.
b) Greater magnification from region 1 in Fig.36a. Note presence of pit/small crack.

63

15558/6/06
Copyright © 2006, TWI Ltd
 1

crack face

a)

crack face

b)

Fig.37 Scanning electron images showing the fracture face of a crack in plate 2205-04.
Magnifications are shown by nominal scales.

a) Overview, backscattered electron image.

b) Greater magnification of area 1 in Fig.37a, showing cleaved (brittle) morphology.

64

15558/6/06
Copyright © 2006, TWI Ltd
 crack face

lab fracture

a)

crack face
1

b)

Fig.38 Scanning electron images showing the fracture face of a crack in plate 316-01.
Magnifications are shown by nominal scales.

a) Overview.
b) As above but at greater magnification.

65

15558/6/06
Copyright © 2006, TWI Ltd
 crack face

c)

lab fracture

d)

Fig.38 Cont’d Scanning electron images showing the fracture face of a crack in plate 316-
01. Magnifications are shown by nominal scales.

c) Greater magnification of area 1 in Fig.38b, showing cleaved (brittle) morphology.

d) Greater magnification of area 2 in Fig.38b, showing dimpled ductile morphology.

66

15558/6/06
Copyright © 2006, TWI Ltd
 crack face
1

lab fracture

a)

b)

Fig.39 Scanning electron images showing the exposed fracture face of the crack shown in
Fig.32c (Pipe 1). Magnifications are shown by nominal scales.

a) Overview.
b) Greater magnification of area 1 in Fig.39a, showing cleaved (brittle) morphology.
67

15558/6/06
Copyright © 2006, TWI Ltd
 A TOP Plate 2205-3

2
1 Scan 5

4 3

Scan 10

5
6
8
7

11 Scan 15

10
15 12
13
14
16

20
17
18
Scan 20
19

21

Scan 25

BOTTOM B
Fig.40 MPI Indications on Plate 2205-02. Scan numbers refer to Eddy current scans.

68

15558/6/06
Copyright © 2006, TWI Ltd
 15

14

13

12

11

10
ACPD readings (mm)

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Crack face length (mm)

Fig.41 Comparison of crack sizes and ACPD readings.

Fig.42 CVM report showing false calls.

69

15558/6/06
Copyright © 2006, TWI Ltd
Fig.43 CVM report showing increase in pressure because of crack.

Fig.44 Example of Eddy current scan on plate 316L-01 (Scan 23). (Flaw in Fig.18b).

Crack 1 (depth=1.2mm)

Crack 2 (depth=2.9mm)

Fig.45 Example of Eddy current scan results on plate 2205-05 (the crack depths are ACPD
results).

70

15558/6/06
Copyright © 2006, TWI Ltd
 Scan with Probe 102P1 (70mm Scan with Weldscan Probe MPI in the same area
from edge) 800PO1 (70mm from edge)

Fig.46 Effect of probe type on signal from a group of cracks (Plate 2205-05).

150

140

130

120

110

100

90
Amplitude

80

70

60
−♦− : 102P01
50
−■− : 801P15L
−*− : 800P01
40

30

20 −×− : 800P01JD1P
10 −▲− : 801P02
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Lift off (mm)

Fig.47 Probe sensitivity against lift off (duplex plate 2205-05, Scan 24).

100
P102P1
P122P1D
90
Series3

80

70
Eddy current readings

60

50

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Crack face length (mm)

a)
Fig.48 Comparison between all the eddy current probes for signal size against crack face
length measured from the metallographic sections on plate 2205-05.

a) Data from high frequency (2MHz) probes.

71

15558/6/06
Copyright © 2006, TWI Ltd
 100
800P01
800P04
90 800P15L

80

70
Eddy current readings

60

50

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Crack face length (mm)

b)
100
800P04, 100kHz
800P04, 300kHz
90
800P04, 600kHz

80

70
Eddy current readings

60

50

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Crack face length (mm)

c)

Fig.48 Cont’d Comparison between all the Eddy current probes for signal size against
crack face length measured from the metallographic sections on plate 2205-05

b) Data from low frequency (100kHz) probes.

c) Data from one probe (800P04) at different frequency (100kHz) probes.

72

15558/6/06
Copyright © 2006, TWI Ltd
 Weld toe
Crack

No crack: Signal of weld toe Crack

Fig.49 Scans using Weldscan probe adapted for hot use (Fig.14) on hot pipe.

Scan1, x*=160mm Scan2, x=169mm Scan3, x=178mm

(*: distance from the side A of the plate)

Fig.50 Scan of Duplex plate 2205-02 (multiple cracking) with ACFM Mini Probe.

−−−Bx: 1st detector

−−−Bx: 2nd detector
−−−Bx: 3rd detector

−−−Bz: 1st detector

−−−Bz: 2nd detector
−−−Bz: 3rd detector

Fig.51 Scan of same plate as Fig 38 with Weld Array probe.

73

15558/6/06
Copyright © 2006, TWI Ltd
Fig.52 ACFM scan of Pipe 1 crack with no lift-off (mini probe).

Fig.53 Scan of same area as above with 2mm lift-off.

74

15558/6/06
Copyright © 2006, TWI Ltd
Fig.54 Scan of hot duplex plate (2205-05) with cooled ACFM Mini Probe (2mm lift off).

Fig.55 Phased array detection of small cracked area in 316L stainless plate 316L-01.

75

15558/6/06
Copyright © 2006, TWI Ltd
Fig.56 Phased array UT detection of multiple cracks on Duplex Plate 2205-02.

Fig.57 Phased array UT detection of crack in Pipe 1.

76

15558/6/06
Copyright © 2006, TWI Ltd
Fig.58 Scan from Eddy current array probe on 316L stainless steel plate.

77

15558/6/06
Copyright © 2006, TWI Ltd
Fig.48

Fig.59 Scan from Eddy current array probe on duplex stainless steel.

-3
x 10
4

) 3
br
a(
e 2
ul
a
V
k 1
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

Fig.60 Typical signals from the pulsed eddy current system (Duplex plate 2205-02).

78

15558/6/06
Copyright © 2006, TWI Ltd
 1

Fig.61 Image from Laser Shearography system (Plate 2205-03).

79

15558/6/06
Copyright © 2006, TWI Ltd
 APPENDIX A
Weld Specifications

15558/6/06
Copyright © 2006, TWI Ltd
 APPENDIX B
Results From the DP, MPI and ACPD of Cracks

15558/6/06
Copyright © 2006, TWI Ltd
 A TOP Plate 2205-2

Scan 5

Crack 1

Scan 10

Crack 2
Crack 3

Scan 15

Crack 4

Crack 5

Crack 6
Scan 20
Crack 7
Crack 8

Scan 25

Scan 30
BOTTOM B
Figure B1: MPI results on duplex plate 2205-2

B1

15558/6/06
Copyright © 2006, TWI Ltd
 Defect Length (mm) Scan number ACPD readings (mm)

8 2.7
1 15 9 2.6
10 3.4
10 3.5
11 1.6
2 19
12 0.8
13 0.1
11 2.6
12 3
Plate 2205-2

13 4
3 36
15 5.1
16 6.5
17 3.5

4 30 17 0

5 50 18.5 0

6 25 18.5 0

7 50 18.5 0.1

20 2.3
21 1.8
8 22
22 3.5
23 0.2

Table B1: ACPD results for some cracks detected on the plate 2205-2

B2

15558/6/06
Copyright © 2006, TWI Ltd
 A TOP Plate 2205-3

2
1 Scan 5

4 3

Scan 10

5
6
8
7

11 Scan 15

10
15 12
13
14
16

20
17
18
Scan 20
19

21

Scan 25

BOTTOM B
Figure B2 : MPI results on duplex plate 2005-3

B3

15558/6/06
Copyright © 2006, TWI Ltd
 Defect Length (mm) Scan number ACPD readings (mm)

6 1.1
1 12 6.5 2.1
7 2.6
6 1
2 10 6.5 2.4
7 3.6
8 4.6
9 5
3 24
10 3.8
11 1.3
Plate 2205-3

13 1.4
5 10
13.5 1.8

6 4 13.5 0.2

14 1.3
9 12
15 0.8

10 8 14.5 0.4

17.5 1.8
12 8
18 1.4

19 4 19 0.5

22 2.4
23 6.5
21 27 24 6.9
25 3
26 1.3

Table B2: ACPD results for some cracks detected on plate 2205-3

B4

15558/6/06
Copyright © 2006, TWI Ltd
 A TOP Plate 2205-4

Crack 1

Scan 5
Crack 3
Crack 2

Crack 4 Scan 10

Crack 5

Crack 6

Crack 7

Scan 15
Crack 8

Crack 9
Crack 8

Scan 20

Crack 10

Scan 25

Crack 11 Crack 12

BOTTOM B
Figure B3: MPI results on duplex plate 2005-4

B5

15558/6/06
Copyright © 2006, TWI Ltd
 Defect Length (mm) Scan number ACPD readings (mm)

1 6 7.5 0.4

2 6 8 0.7

7 0.7
3 10 8 1.3
9 0.3

4 5 8.5 0
Plate 2205-4

5 5 8.5 0.5

6 5 9 0

14 1.7
7 16 15 2.1
16 0.3

8 2 15 0

19 1.3
20 5
9 22
21 4.8
22 1.2
23 0.7
10 16 24 2.5
25 1.3

11 3 24.5 0

12 2 25.5 0.2

Table B3: ACPD results for some cracks detected on plate 2205-4

B6

15558/6/06
Copyright © 2006, TWI Ltd
 A TOP Plate 2205-5
Crack 1 Crack 2

Scan 5
Crack 3
Crack 4
Crack 5

Crack 6 Crack 7
Crack 8

Scan 10

Crack 10 Crack 9

Scan 15
Crack 11

Crack 12
Crack 13

Crack 14 Scan 20

Scan 25

BOTTOM
Figure B4: MPI results on duplex plate 2005-5 B
B7

15558/6/06
Copyright © 2006, TWI Ltd
 Defect Length (mm) Scan number ACPD readings (mm)

5 1.2
1 11
6 2
5 2.9
2 10
6 3.9
7 12.4
11 7.7
13 9.3
15 23.4
17 26.6
3 105
19 10.3
22 7.5
25 5.2
27 1

7 4
4 16 8 5
Plate 2205-5

9 4.7
10 4.6
11 4.4
8 24 12 3.7
14 0.7
14 6.5
15 17.5
9 28 16 23.5
18 3.4
14 1.7
15 2.4
10 35 16 1.9
17 3.4
18 3.5
17 1.4
19 4.1
11 45 21 1.6
23 5.1
25 1.3
20 3.1
12 23 21 2.5
22 0.8
22 2.6
14 15 23 3.8

Table B4: ACPD results for some cracks detected on plate 2205-5

B8

15558/6/06
Copyright © 2006, TWI Ltd
 B BOTTOM Plate 2205-6

Crack 1

Crack 3
Scan 25
Crack 2

Crack 7

Crack 4

Crack 8
Scan 20
Crack 6

Crack 10

Crack 11
Crack 9

Crack 15 Scan 15
Crack 12

Crack 13

Crack 16
Crack 14
Crack 18 Scan 10

Crack 17

Crack 21

Crack 19 Crack 20

Scan 5

TOP A
Figure B5: MPI results on duplex plate 2005-6

B9

15558/6/06
Copyright © 2006, TWI Ltd
 Defect Length (mm) Scan number ACPD readings (mm)

1 5 25.5 0

2 1 24.5 0

3 1 24.5 0

4 1 24.5 0

20 1.6
6 17 19 2
18 1.7
21 0.3
8 10
20 0.2
Plate 2205-6

16 0.6
12 10
15 0.8

13 1 13.5 0

14 1 13 0

15 5 13 0.6

10 1.3
16 16 9 1.3
8 0.8

17 2 9 0.2

18 5 9 0.1

19 8 9 0.2

20 5 10 0.2

21 3 9 0.5

Table B5: ACPD results for some cracks detected on plate 2205-6

B10

15558/6/06
Copyright © 2006, TWI Ltd
Figure B6 PT results on the 316L plates

Defect 1

Defect 2

Figure B7 PT results on the 316L plates

Defect 1 Max ACPD = 0.8mm

Defect 2 Max ACPD= 1.1mm

B11

15558/6/06
Copyright © 2006, TWI Ltd
Figure B8 PT results on the plate 316L-W1

Figure B9 MPI results on the Pipe 1

Crack in pipe #1
ACPD readings

3
2
(mm)

1
0
-25 -20 -15 -10 -5 0 5 10 15 20
Distance from TDC

Figure B10 ACPD results on the Pipe 1

B12

15558/6/06
Copyright © 2006, TWI Ltd
Figure B11 MPI results on the Pipe 2

Crack in pipe #2
ACPD readings

3
2
(mm)

1
0
-10 -5 0 5 10 15 20
Distance from TDC (mm)

Figure B12 ACPD results on the Pipe 2

B13

15558/6/06
Copyright © 2006, TWI Ltd
 APPENDIX C
Acoustic Emission Tests

15558/6/06
Copyright © 2006, TWI Ltd
 ABERDEEN OIL & GAS OPERATIONS
TITLE: Acoustic Monitoring of 316L and Duplex Stainless Steel
Pipeline

CLIENT: TWI
REPORT No: O-1904
DATE: 11th April 2006

© Lloyd’s Register EMEA, 2006

All rights reserved
 Technical Report Document Control Page
1. Report No. 2. Report date 3. Revision 4. Comments
O-1904 11/04/06 A

5. Title & Subtitle 6. Security classification of this report

Acoustic Monitoring of 316L and Duplex Stainless Steel Restricted to LR and Client
Pipeline

7. Author (name & signature): 8. Security classification of this page

Restricted to LR and Client

Alan Smith
9. Reporting organisation name and address 10. Reporting organisation reference(s)
Lloyd's Register EMEA
25 Union Terrace, Aberdeen, AB10 1NN

11. Sponsoring organisation name and address 12. Sponsoring organisation reference(s)
TWI
Granta Park,
Abington,
Cambridge CB1 6AL

13. Summary
Test specimens of type 316L Austenitic plate, 22% Cr Duplex Stainless Steel plate and also two duplex stainless
pipelines with a girth weld in the middle, were loaded in bending, heated to 150°C and exposed to a saline
environment with the intention of initiating stress corrosion cracking. Several 600kHz acoustic emission sensors
were installed and the specimens were monitored for evidence of chloride stress corrosion cracking. The report
details the results of the monitoring exercise.
14. Key words 15. Distribution statement
CSCC, Acoustic Emission, Duplex stainless steel

16. Checked by (name & signature): 17. Authorised by (name & signature):

Lloyd’s Register, its affiliates and subsidiaries and their respective officers, employees or agents are, individually
and collectively, referred to in this clause as the ‘Lloyd's Register Group’. The Lloyd's Register Group assumes no
responsibility and shall not be liable to any person for any loss, damage or expense caused by reliance on the
information or advice in this document or howsoever provided, unless that person has signed a contract with the
relevant Lloyd's Register Group entity for the provision of this information or advice and in that case any
responsibility or liability is exclusively on the terms and conditions set out in that contract.

© Lloyd’s Register EMEA, 2006

All rights reserved
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI
EXECUTIVE SUMMARY
Chloride stress corrosion cracking (CSCC) was induced in a series of test specimens made
from 22% Chrome Duplex stainless steel and type 316 austenitic stainless steel. Many of
these specimens were continuously monitored throughout the test period using an acoustic
emission data acquisition unit developed by a consortium of Lloyd's Register, Airbus
Industries and Ultra Electronics. The equipment used was a prototype version of a
commercially available unit marketed under the title Vigilant. From the series of assessments
the most significant samples were two pipe specimens made of 22% Cr Duplex stainless steel.
This report describes the pipe tests in detail.
The two pipes were monitored with a series of resonant sensors collectively covering a
bandwidth of frequencies between 100kHz to 800kHz. It was found that CSCC could most
clearly be identified using sensors resonating at an optimum frequency range of between
450kHz and 800kHz. However, there was a marked difference between the two tests with
which CSCC could be identified from the interpretation of the acoustic emission signals. In
the first specimen CSCC could only be identified by focussing attention on a particular
section of the pipe known to be associated with stress corrosion cracking. For the second
specimen identification was easier and a cluster of activity could be discriminated directly
from the ever present background data.
Further analysis of the acoustic emission signals showed that stress corrosion cracking was
characterised by an increase in signal amplitude with respect to the background noise levels
when corrected for signal attenuation. This provided a useful means of discriminating
between stress corrosion cracking and other sources of acoustic emissions. However, other
means of identifying stress corrosion cracking from the acoustic emission spectra was
possible but these were less reliable. It should be noted that the tests were carried out under
relatively quiet laboratory conditions compared to a typical industrial location with greater
acoustic background noise spectra and that discrimination of stress corrosion cracking in
some working environments will thus prove more difficult.

Reference: O-1904, Rev A Page i of i

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

TABLE OF CONTENTS
1 GLOSSARY.................................................................................................................................... 1
2 INTRODUCTION .......................................................................................................................... 3
2.1 Background ............................................................................................................................ 3
3 EXPERIMENTAL PROCEDURE ................................................................................................. 4
3.1.1 General ..................................................................................................................... 4
3.1.2 22% Cr Duplex Stainless Steel Pipes....................................................................... 4
3.2 Acoustic emission monitoring configuration ......................................................................... 5
3.2.1 Monitoring Equipment: Specimen DP 1 ................................................................... 5
3.2.2 Monitoring Equipment: Specimen DP2 .................................................................... 5
3.2.3 BALRUE DAU configuration.................................................................................... 5
3.2.4 Guard Sensor Configuration..................................................................................... 6
3.2.5 Analysis Software...................................................................................................... 6
4 RESULTS ....................................................................................................................................... 7
4.1 General................................................................................................................................... 7
4.2 Specimen DP1 (600kHz response only) ................................................................................ 7
4.3 Specimen DP2 (Wide band response).................................................................................... 8
5 DISCUSSION ................................................................................................................................. 9
6 CONCLUSIONS........................................................................................................................... 11
7. REFERENCES.............................................................................................................................. 12
8 ACKNOWLEDGEMENT ............................................................................................................ 12
APPENDIX A APPENDIX 1.......................................................................................................... 29

Reference: O-1904, Rev A Page ii of i

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

1 GLOSSARY
Term Definition
Acoustic The energy associated with an acoustic burst measured from the response of
Power the sensor
AE The amplitude of the burst emission, measured in dB scale referred to the
amplitude preamplifier input, Vr = 10-6V, AE dB = 20 log10(V/Vr)
Amplitude A threshold level of amplitude set in the DAU for the first three sensors to be
level hit. (Signals of amplitude less than this are not recorded)
Array A group of sensors configured to monitor a certain region
Backgound The general level of signal recorded when no discrete bursts are measured
noise (also known as average signal level (ASL)
Burst Also referred to as event, AE signal having an identifiable beginning and end
Cluster An AE source identifiable spatially in the geometric or ∆T space domain
DAU Data acquisition unit
Filter Process of classifying data according to chosen parameters
Grain size The dimensions of the ∆T matrix cell within which the burst must be recorded
to be measured as significant (the resolution is half the grain size)
Guard sensor An AE sensor used to discriminate sources originating outside the area of
interest
Hit The detection of one burst on a sensor (hit order refers to the sequence of
sensors hit by an individual burst)
Hotspot A cluster identifiable on the ∆T plot as having a high number of hits with
similar time difference.
Hsu-Nielsen Device to simulate an AE event using the fracture of a brittle graphite pencil
source lead
K Detectability, the difference between the AE detection threshold (ASL+4dB)
and the attenuation curve at the last hit sensor (measured in dB).
Leading edge The use of the wave front to trigger the timers used to identify the bust in the
time domain (i.e. when the burst first exceeds the detection threshold)
CSCC Chloride Stress corrosion cracking, a failure mechanism triggered by the
conjoint action of stress and corrosion of the passivating layer in the presence
of Chloride ions.
Tick time Time interval set on the DAU during which the key features of the burst are
recorded
∆T The difference between the arrival times of a burst signal at two sensors

Reference: O-1904, Rev A Page 1 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

∆T plot A diagram showing the time difference between two pairs of sensors
Wait time Time duration set on the DAU which must be exceeded for two events to be
regarded as discrete

Reference: O-1904, Rev A Page 2 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

2 INTRODUCTION
2.1 Background
Acoustic emissions are ultrasonic stress waves (pulses of sound) produced during processes
such as crack growth in metals. They can originate when the metal grains fracture, releasing
strain energy which propagates through the metal as compressive, shear and plate waves. The
aim of acoustic emission monitoring is to detect these stress waves and translate the wave
energy into electrical signals. By placing a network of suitable sensors around a region of
interest and recording key parameters (such as the time difference between the signal hitting
different sensors) the location of the acoustic emission can be identified together with
pertinent information relating to the signal strength etc.
The BALRUE system of acoustic monitoring (developed by Lloyd's Register, Airbus
Industries and Ultra Electronics) uses a sophisticated signal conditioning algorithm to identify
acoustic signals by the time difference between the signal hitting successive sensor pairs. A
time domain matrix is created and signals received with similar time differences (∆T) are
allocated a volumetric cell within the component cell matrix. During stress corrosion cracking
it was anticipated that the signals would be repeatedly received with similar ∆Ts. The system
software was programmed to recognise such patterns and record these (together with their
associated parameters) internally. In this way, random signals can be identified and
disregarded. Thus data records are censored to retain significant information. Data records are
therefore small and manageable, allowing data downloads using conventional data transfer
methods such as a modem or email. The data signals are then post processed to assess their
significance. This system is configured such that it requires that an individual acoustic
emission burst to be detected by at least three sensors before the acoustic noise burst is
processed further.

Reference: O-1904, Rev A Page 3 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

3 EXPERIMENTAL PROCEDURE
3.1.1 General
The test program was divided into a series of tests comprising:
• Unwelded plate specimens of type 316 stainless steel
• Unwelded plate specimens of 22% Chrome Duplex
• Welded plate specimens of type 316 stainless steel
• Welded pipe specimens of 22% Chrome Duplex
This report concentrates on the acoustic monitoring of the 22% Cr Duplex welded pipe test
specimen results. Results for the plate trials are presented in appendix A. The installation of
the acoustic monitoring system for the 22% Cr Duplex pipe specimens was carefully
controlled by Lloyd's Register whereas the other tests were set up and monitored by TWI
under less direct control of Lloyd's Register. The acoustic monitoring of the non welded tests
was originally intended as a control measure only to assess the typical background noise
levels and configuration parameters to establish accurate testing conditions and hence limited
importance was attached to the plate tests. In addition the small size of the welded test plates
made it extremely difficult to discriminate individual bursts from signals reflected from the
specimen sides. Thus the plate trials did not realistically reflect the capabilities of the acoustic
monitoring equipment in typical field applications..

3.1.2 22% Cr Duplex Stainless Steel Pipes

The test sample configuration consisted of two duplex stainless steel pipes joined with a girth
weld, was loaded in a test frame using four point bending (fig 1). Resistance heating blankets
were then draped over the pipes as shown. Standard resonant type acoustic sensors were
attached at the positions shown (Fig 1 and on the pipe surface development plans, Fig 3 and
11) with calibration checks and attenuation measurements being made. The welded pipe was
then heated to 150°C at the weld region (typically 130°C at the position of the nearest
acoustic sensor), loaded in four point bending to induce yield level stresses near the pipe
surface. This was then left for three days. During the initial three day period the pipes were
monitored using the acoustic emission equipment (details given below). On the third day a
saline solution was dripped onto the top of the pipe surface near the central weld. There were
a total of two test samples and these specimens was acoustically monitored and subjected to a
hot dye-penetrant test at approximately daily intervals. The tests were stopped at the first sign
of cracking identified from these dye-penetrant tests.

The two tests differed in the detail of the acoustic emission system applied. Following the first
test (DP1) it was decided that additional useful information could be supplied by extending
the frequency bandwidth of the ultrasonic sensors from 400kHz to 800kHz (as used in the
first pipe test) to 100kHz to 800kHz. This was done by applying additional sensors. Details of
each pipe sample test configuration are described below.

Reference: O-1904, Rev A Page 4 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

3.2 Acoustic emission monitoring configuration

3.2.1 Monitoring Equipment: Specimen DP 1

The acoustic emission system comprised of a 24 channel standard BALRUE Data Acquisition
System (DAS), with selected sensors and matching preamplifiers. A coaxial conductor
supplied power from the DAS to the sensors via a matched preamplifier and transmitted the
acoustic data back from the sensors to the DAS via the preamplifiers. Six, type NA3303
600kHz sensors were installed to each pipe using resonant characteristics optimised for use
with 400kHz to 800kHz band pass preamplifiers. The preamplifiers were type PA3301
600kHz with a gain of 40dB and optimised for resonance between 400kHz and 800kHz.
The sensors were attached at the positions shown in Fig 1. This comprised of 4 sensors
attached as asymmetric pairs either side of the weld. An additional 2 guard sensors were
attached at the supports and these were configured to disregard any noise hitting the support
sensors first. The response of individual sensors to a 0.3mm Hsu-Nielsen source was
measured together with the ambient noise level (e.g. see Ref 1). Signals were generally in
excess of 95dB amplitude with many sensors recording 102.3dB (i.e. the sensor signal was at
saturation). Readings of the ambient noise levels were taken together with an attenuation plot
for the breaking of a 0.3mm Hsu Nielsen source (see Table 1 to 4). It can be seen (Fig 3 and
11) that the detectability parameter, K, corresponding to the sensor spacing used in the
experimental configuration exceeded 40dB.
Additional measurements were used to confirm the accuracy of the Tobias algorithm in
locating sources (Ref 2).

3.2.2 Monitoring Equipment: Specimen DP2

Specimen DP2 was set up as described for specimen DP1 (above). However, additional
sensors were applied to increase the frequency bandwidth sensor response. Thus, array 2 was
set up using four, 300kHz sensors with a gain of 40dB and band pass characteristics optimised
for resonance between 200kHz and 400kHz. Array 3 was set up using four, 150kHz sensors
with a gain of 40dB and band pass characteristics optimised for resonance between 100kHz
and 200kHz. The sensor layout is shown in Fig 1 and 11.

3.2.3 BALRUE DAU configuration

The BALRUE DAU was configured using the programme HCI cracks (Ref 3) as shown in
Table 2. This configuration was set up to with a low cluster threshold i.e. only 2 signals had to
be received with a similar time difference (corresponding to a particular location) for the
source to be identified as significant. The deviation allowed in the time difference for a
particular volume matrix cell was set to +/-2µs. This was deliberately established as being
fine as this corresponds to about 10mm at the centre of pipe. The amplitude threshold was set
to reject signals obtained from background noise, whilst retaining enough sensitivity to retain
significant data should a crack initiate. The system was designed to respond to an incoming
signal and remain dormant when no signals are received. Data was recorded using the leading
edge mode.

Reference: O-1904, Rev A Page 5 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

3.2.4 Guard Sensor configuration

Sensors 5 and 6 were attached to the central supports and were configured to guard against
any ambient noise from the pipe supports for specimen DP1. Two 600kHz sensors were
attached to the central supports for specimen DP2 but were not configured into the test as
censoring of the data for this test was done using the filtering algorithm at the post processing
stage.

3.2.5 Analysis Software

Post process signal analysis was done using the programme BALRUE Data Analysis
Environment, Version 1.0.06.

Reference: O-1904, Rev A Page 6 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

4 RESULTS
4.1 General
The following processes are used when analysing acoustic data.
• Filtering. The data is sorted according to parameters such as date, array (effectively
sensor type) and signal amplitude etc
• Clustering. This is used to group and rank the most commonly occurring ∆T values
• ∆T plotting. The data from the most commonly occurring clusters is displayed in a
graphical form from which an area of intense acoustic activity or "hot spot" can be
identified and isolated. Supplementary filtering based on ∆T coordinates associated
with the hot spot can then be used to identify salient features associated with CSCC.
• Locate. The Tobias algorithm is used to locate the AE source in real space.
• Attenuation comparison. The signal amplitude as the signal reaches successive sensors
is recorded for comparison against the original attenuation curve
To assist in interpretation, data was filtered by sensor frequency and ∆T coordinates. This
filtering then isolated acoustic activity located near the weld which was from activity
emanating from under the heating pad.

4.2 Specimen DP1 (600kHz response only)

A significant amount of data was recorded before the saline solution was dripped on the pipe
surface. Much of this anomalous data was largely consistent with bursts detected close to
sensor 4 or outside of the sensor array. Such data was consistent with background noise from
the horns of the support saddle. This data was therefore disregarded as anomalous. In order to
interpret the acoustic data meaningfully it was therefore necessary to filter out all data that
"hit" sensors 4, 5 and 6 first. Initial analysis of the remaining data showed the events to be
very dispersed with no strong acoustic "hot spot" (Fig 2 and 3)
Further analysis concentrated on the region near to (within approximately 50mm radius) the
12 o'clock position of the central girth weld. The geometric distribution of data is shown in
Fig 4.
Relatively few events (6) emanating from this central region exceeded the established
recording thresholds set up. The events which did exceed the reporting threshold were all
recorded on the 6th January (see Fig 5). It was noted that the background noise levels of the
sensors remained reasonably constant during this period (Fig 6)
Comparison of the signals from the central region with the attenuation curve show that the
amplitude is between 10dB and 30dB of the calibration curve (see Fig 7).
To place the above in context, a histogram of all the events (excluding those that hit sensor 4
first and the data from the suspected CSCC source) is shown for comparison in Fig 8. It can
be seen that the distribution of amplitudes for the CSCC data is more heavily weighted
towards higher energy signals than the general data, but it should be noted that the source data
set is restricted to about 12 events.

Reference: O-1904, Rev A Page 7 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

4.3 Specimen DP2 (Wide band response)

Specimen DP2 used additional sensors to enable a comparison to be made with signals
received from sensors tuned to a lower resonant frequency. The number of acoustic events
recorded for specimen DP2 was significantly higher than for specimen DP1. Acoustic
"hotpots" were visible using either the 600kHz, 300kHz and 150kHz sensors and the hotspots
correlated well with the anticipated site of the CSCC damage. From comparison of Fig 8 to 9,
15 to 16, 19 to 20 and 23 to 24, it can be seen that the amplitude of the events recorded by the
600kHz sensors were more strongly weighted towards the higher amplitude signals than the
signals recorded by the 300kHz and 150kHz sensors and that the amplitudes of signals
detected with the 600kHz sensors was most easily differentiated from the ambient noise levels.
The 600kHz sensors were therefore the most suited to monitoring CSCC in this material
within the overall bandwidth investigated.
Looking at the results from the 600kHz sensors in more detail (Figs 12 to 14) it can be seen
that the first signs of significant acoustic activity were detected on 9 January at 21:30hrs and a
significant jump in activity was seen later on 11 January at about 6:00am. The difference
between the source amplitude and the background noise level was greater than 20dB.

Reference: O-1904, Rev A Page 8 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

5 DISCUSSION
Based on discussion and other practical experiences, the following sentencing criteria have
been proposed as an aid to diagnosis of CSCC:
• Total burst count recorded within a given ∆T matrix cell and burst count trend within
a given cluster.
• Burst cluster ratio defined as highest number of events recorded within a matrix cell
associated with a source divided by number of events recorded within 20th highest
matrix cell associated with the same source.
• Comparison of signal amplitude with background noise level
• Source location with respect to likely sources (e.g. welds)
• Signal amplitude compared to Hsu-Nielsen attenuation curve (including attenuation)
Over and above these issues, it is important to recognise that the ability of any AE system to
effectively identify CSCC is dependent on the initial system configuration. The configuration
parameters for the above tests were deliberately set to allow as much data through as possible
without compromising the effectiveness of the system. This discussion will examine the
above sentencing guidelines and compare these to the experimental results to identify those
criteria key to the diagnosis of CSCC in duplex stainless steels.
The first point of interest is the extreme rapidity with which CSCC appeared to start from the
saline drips being applied. This highlights the potential difficulty in applying any form of
inspection other than continuous monitoring if CSCC is to be safeguarded against. Of the two
pipe tests, the acoustic data obtained from specimen DP1 showed very subtle changes
compared to the background noise, whereas specimen DP2 gave a clearer signal which could
be more easily discriminated. This lack of consistency was also a feature common to the
preliminary plate tests, although the application of the acoustic emission to the plate tests was
less rigorously supervised and therefore the results have been formally excluded from this
analysis. It is difficult to say why the system worked well for some pipe and plate tests and
not so well for others. Great care was taken to apply the system, test the response of the
system and conduct the test. The only significant difference between the two pipe tests (other
than the application of additional sensors to DP2) was the integration of the guard sensors on
DP1 into the DAU configuration software. However, the location of the guard sensors was
such that no bursts generated by CSCC were excluded. The most probable reason for the
inconsistency between the two tests is that the acoustic response of the Duplex parent material
to CSCC is itself inconsistent.
The burst count (measured cumulatively) associated with CSCC in specimen DP1 was very
low (12 bursts only were recorded). The burst count was clearly dependent on the amplitude
level with which the monitoring equipment is configured and the background noise levels.
Setting these parameters lower would have increased the burst count significantly but it would
also have increased the ambient signals filtering through to the system. The setting of the first
hit amplitude level to approximately 30dB below the attenuation curve provided a realistic
compromise between adequate detection capability and data management constraints. For the
reliable setting of cumulative burst count as a sentencing criteria a very low limit (less than 12)
would need to be set together with a coarse matrix resolution. This would prove unworkable
in many working environments as ambient signal levels are quite likely to exceed this, unless
the monitored region can be sufficiently confined to minimise ambient effects or the

Reference: O-1904, Rev A Page 9 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

environment is exceptionally quiet. Looking at the data from specimen DP2, setting the
cumulative burst count criteria to a given value (say 50) for the individual matrix cell would
have resulted in the criteria being triggered if the resolution of the matrix cell was extremely
coarse.
The trend of the cumulative burst count does provide a clue to the onset of CSCC as can be
seen (Fig 5 and 12). A characteristic of CSCC is that growth is rapid and step like. A possible
variation of this criteria would be look at the number of bursts emitted within a given time
frame In consideration of this however, the number of events recorded for DP1 was so few
that there is little to distinguish this from the ambient burst rate. Like wise there was little to
distinguish the acoustic power response (or comparison of acoustic power to cumulative burst
count) from ambient levels for either specimen.
The burst cluster ratio was negligibly low for the test of specimen DP1. Specimen DP2 had a
burst cluster threshold ratio of 3.75. It follows that setting the burst cluster threshold to greater
than 3 would result in some cases of CSCC going undetected. Even at a cluster threshold of 3
practical experience shows that many events attributable to sources other than CSCC would
filter through. Hence discriminating by the cluster threshold is a relative poor means of
diagnosing CSCC.
The signal amplitude was clearly distinguished from the background noise during the test.
This test is not intended as a criteria that must be satisfied in order to diagnose CSCC but
rather to identify possible other causes rather than CSCC (e.g. friction) and alert the analyser
to possible spurious data.
The source location was reasonably good and correlated well with the known location of the
CSCC. It is a valid criteria for the diagnosis of CSCC. The location was then used to compare
the signal amplitude to the attenuation curve. In both cases the data associated with CSCC
was more heavily weighted to higher amplitudes than the ambient data. As such this
comparison forms one of the most reliable guides to the detection of CSCC. A suitable simple
criterion based on signal amplitude is therefore that, signals within 15dB of the attenuation
curve together with a shift in the recorded amplitude of the suspected source with respect to
the ambient data, should be regarded as indicative of the possible presence of CSCC.
If the test is used in the local mode a specific region can be identified such that a higher
probability of detection can be ensured by setting configuration parameters so that any signals
(but particularly high amplitude signals within 25dB of the attenuation curve) trigger an alarm
level at very low burst counts. This would lead to some false calls and the chance of a false
call would increase with the size of the monitored area and the level of background noise. The
equipment can also be used in a global mode whereby high amplitude events recorded in
relatively large numbers would be required to alert an alarm level. Global surveillance would
be at the expense of possibly missing some instances of cracking but would be more
appropriate to applications which were more defect tolerant. It should be noted that the above
trials were conducted in relatively benign acoustic environmental conditions. When assessing
the effectiveness of acoustic emission in practice, actual levels of ambient noise may be much
higher. The trials illustrate the importance of setting up representative trials to determine
effective assessment criteria when using acoustic emission as a diagnostic tool.

Reference: O-1904, Rev A Page 10 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

6 CONCLUSIONS

The two welded pipe 22% Duplex pipe specimens were loaded using four point bending and
subjected to a saline solution drip applied at elevated temperature in order to induce chloride
stress corrosion cracking. During loading both pipes were monitored for acoustic emission
using the Balrue acoustic emission data analysis unit (marketed as "Vigilant"). The following
conclusions were drawn:
• Significant levels of acoustic emission were detected for specimen DP1 on the 6
December 2005 and for specimen DP2 on 9 January 2006.
• The acoustic response of specimen DP2 to the assessed onset of CSCC was much
clearer than the response of specimen DP1.
• The acoustic signal in the bandwidth 450kHz to 800kHz from both specimens
emanating from CSCC were associated with higher amplitudes than the prevailing
ambient noise when corrected for signal attenuation.
• Of three types of sensors tuned to provide a resonant response at 150kHz, 300kHz and
600kHz, the sensors tuned to 600kHz gave the clearest response to CSCC in
comparison to ambient levels of background noise.
• Monitoring using acoustic emission to detect CSCC can either be done by setting low
thresholds at the expense of increasing the number of spurious calls, or by setting
higher thresholds to eliminate spurious calls at the expense of possibly missing regions
of CSCC.

Reference: O-1904, Rev A Page 11 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

7 REFERENCES
[1] CEN/TC 138/WG7- Non destructive testing- Acoustic Emission- General principles
January 1999
[2] A Tobias, Acoustic emission source location in two dimensions by an array of three and
four sensors, Central Electricity Generating Board, January 1975 .
[3] BALRUE Data Analysis Environment, Version 1.0.06 developed jointly by Lloyd's
Register, Airbus Industries and Ultra Electronics, obtainable from Ultra Electronics,
Weymouth, Dorset, UK.

Reference: O-1904, Rev A Page 12 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

8. ACKNOWLEDGEMENT

The author would like to acknowledge the assistance given by Dr LM Rogers in this project
and for the development of the BALRUE DAU monitoring unit.

Reference: O-1904, Rev A Page 13 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Leading edge settings

Grain size, µs 4

Resolution, µs +/-2

Cluster threshold 2

Maximum ∆T, µs 400

Amplitude threshold levels for 1st, 2nd and 3rd hit 65:55:45
sensors, dB
Major threshold level, dB 80

Wait time, µs 0.5

Tick time, µs 7.81

Sensor No Average signal Noise

level, dB threshold
level
1 38 42
2 39 43
3 44 48
4 36 40
5 37 41
6 44 48

Table 1 BALRUE DAS parameters for test DP1 (Array 1: 600kHz sensors)

Reference: O-1904, Rev A Page 14 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Leading edge settings

Grain size, µs 4

Resolution, µs +/-2

Cluster threshold 2

Maximum ∆T, µs 400

Amplitude threshold levels for 1st, 2nd and 3rd hit 65:55:45
sensors, dB
Major threshold level, dB 80

Wait time, µs 0.5

Tick time, µs 7.81

Sensor No Average signal Noise

level, dB threshold
level
1 37 41
2 36 40
3 37 41
4 36 40
5 36 40
6 40 44

Table 2 BALRUE DAS parameters for test DP2 (Array 1: 600kHz sensors)

Reference: O-1904, Rev A Page 15 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Leading edge settings

Grain size, µs 4

Resolution, µs +/-2

Cluster threshold 2

Maximum ∆T, µs 400

Amplitude threshold levels for 1st, 2nd and 3rd hit 65:55:45
sensors, dB
Major threshold level, dB 80

Wait time, µs 0.5

Tick time, µs 7.81

Sensor No Average signal Noise

level, dB threshold
level
7 35 39
8 37 41
9 35 39
10 36 40

Table 3 BALRUE DAS parameters for test DP2 (Array 2: 300kHz sensors)

Reference: O-1904, Rev A Page 16 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Leading edge settings

Grain size, µs 4

Resolution, µs +/-2

Cluster threshold 2

Maximum ∆T, µs 400

Amplitude threshold levels for 1st, 2nd and 3rd hit 65:55:45
sensors, dB
Major threshold level, dB 80

Wait time, µs 0.5

Tick time, µs 7.81

Sensor No Average signal Noise

level, dB threshold
level
11 32 36
12 35 39
13 32 36
14 33 37

Table 4 BALRUE DAS parameters for test DP2 (Array 3: 150kHz sensors)

Reference: O-1904, Rev A Page 17 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 1 Experimental setup for specimen DP2

Notes:
1. Sensor arrangement for DP1 was similar but used innermost AE sensors only
2. Innermost sensors 600kHz, middle sensors 300kHz, outer sensors 150kHz

Fig 2 ∆T space diagram for DP1 showing number of hits for associated with time difference
(first hit of 4, 5 or 6 excluded)

Reference: O-1904, Rev A Page 18 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 3 Geometric distribution of acoustic events recorded during test of DP1

Fig 4 Geometric distribution of cluster identified with CSCC for specimen DP1 (red indicates
1-2 bursts)

Reference: O-1904, Rev A Page 19 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 5 Cumulative burst count and cumulative acoustic power associated with CSCC cluster

Fig 6 Amplitude of bursts and related average signal level (ASL) for cluster associated with
CSCC of DP1

Reference: O-1904, Rev A Page 20 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

110

100

90

-0.0407
y = 114.45x
80
amplitude, dB

70 Calibration data
background noise level
SCC data
60 Power law fit to calibration data

50

40

30

20
0 100 200 300 400 500 600 700 800 900
distance from source, mm

Fig 7 Comparison of burst amplitude to attenuation curve for cluster associated with CSCC of
DP1

3
Frequency

2
1
0
10 15 20 25 30 More
dB below HN line

Fig 8 Difference between burst amplitude and attenuation curve corrected for distance for
bursts associated with CSCC for DP1

40
Frequency

30
20
10
0
10 15 20 25 30 More
dB below HN line

Fig 9 Difference between burst amplitude and attenuation curve corrected for distance for all
bursts excluding those with CSCC for DP1

Reference: O-1904, Rev A Page 21 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 10 ∆T space diagram for DP2 showing number of hits for associated with time difference
(600kHz sensors)

Fig 11 Geometric distribution of acoustic events associated with SCC recorded during test of
DP2 (600kHz sensors)

Reference: O-1904, Rev A Page 22 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 12 Cumulative burst count and cumulative acoustic power associated with CSCC cluster
for DP2 (600kHz sensors)

Fig 13 Burst amplitude and average signal level associated with CSCC cluster for DP2
(600kHz sensors)

Reference: O-1904, Rev A Page 23 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

120

100

80
y = -0.0206x + 100.89
Amplitude, dB

600kHz response
150kHz ASL
60
Series7
Linear (600kHz response)

40

20

0
0 100 200 300 400 500 600 700 800 900
Distance from source, mm

Fig 14 Comparison of burst amplitude to attenuation curve for cluster associated with CSCC
of DP2 (600kHz sensors)

60
Frequency

40
20
0
10 15 20 25 30 More
dB below HN line

Fig 15 Difference between burst amplitude and attenuation curve corrected for distance for
bursts associated with CSCC for DP2 (600kHz sensors)

250
Frequency

200
150
100
50
0
10 15 20 25 30 More
dB below HN line

Fig 16 Difference between burst amplitude and attenuation curve corrected for distance for all
bursts associated with DP2 (excluding CSCC cluster), 600kHz sensors

Reference: O-1904, Rev A Page 24 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 17 Distribution of acoustic bursts obtained with 300kHz sensors for specimen DP2
Note: Data cluster at lower left hand corner is considered anomalous

120

100

y = -0.0143x + 101.34
80
Amplitude, dB

300kHz
300kHz ASLl
60
SCC data
Linear (300kHz)

40

20

0
0 100 200 300 400 500 600 700 800 900
Distance from source, mm

Fig 18 Comparison of burst amplitude to attenuation curve for cluster associated with CSCC
of DP2 (300kHz sensors)

Reference: O-1904, Rev A Page 25 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

50
Frequency

40
30
20
10
0
10 15 20 25 30 More
dB below HN line

Fig 19 Difference between burst amplitude and attenuation curve corrected for distance for
bursts associated with CSCC for DP2 (300kHz sensors)

1500
Frequency

1000

500
0
10 15 20 25 30 More
dB below HN line

Fig 20 Difference between burst amplitude and attenuation curve corrected for distance for all
bursts associated with DP2 (300kHz sensors) excluding CSCC cluster

Reference: O-1904, Rev A Page 26 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 21 Distribution of acoustic bursts obtained with 150kHz sensors for specimen DP2
120

100

80
y = -0.0145x + 95.219
Amplitude, dB

150kHz
150kHz ASL
60
Series1
Linear (150kHz)

40

20

0
0 100 200 300 400 500 600 700 800 900
Distance from source, mm

Fig 22 Comparison of burst amplitude to attenuation curve for cluster associated with CSCC
of DP2 (150kHz sensors)

Reference: O-1904, Rev A Page 27 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

15
Frequency

10

0
10 15 20 25 30 More
dB below HN line

Fig 23 Difference between burst amplitude and attenuation curve corrected for distance for
bursts associated with CSCC for DP2 (300kHz sensors)

100
Frequency

80
60
40
20
0
10 15 20 25 30 More
dB below HN line

Fig 24 Difference between burst amplitude and attenuation curve corrected for distance for all
bursts associated with DP2 (150kHz sensors) excluding CSCC cluster

Reference: O-1904, Rev A Page 28 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 Acoustic Monitoring of 316L and Duplex Stainless Steel Pipeline

TWI

Fig 25 Photograph of CSCC in Duplex pipe specimen DP1. Note that the cracking is at the
edge of the heat affected zone boundary with the parent pipe material. (ACPD measured the
cracking to be about 7mm deep, but this remains unconfirmed at the time of writing).

Reference: O-1904, Rev A Page 29 of 29

Date: 11/04/06
File Name: ABERDEEN OIL DRAFT TO TWI Original O-1904.doc
 APPENDIX D
Eddy Current Tests

15558/6/06
Copyright © 2006, TWI Ltd
 Settings: Probe 800P01 (100kHz)
Duplex plates Lift off: 1.18mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 28 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 48.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 115.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A
Duplex plates Lift off: 2.36mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 28 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 54.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 127.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A
Duplex plates Lift off: 3.54mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 29 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
D1

15558/6/06
Copyright © 2006, TWI Ltd
 Ch1 Gain 1G 63.0dB Ch2 Gain 2G 30.0dB
Ch1 Phase 1P 127.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP +20dB Bal. Load LO ---
Graticule GR Rect.A
Duplex plates Lift off: 4.72mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 30 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 70.0dB Ch2 Gain 2G 30.0dB
Ch1 Phase 1P 127.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP +20dB Bal. Load LO ---
Graticule GR Rect.A
Plate: 316-02 Lift off: 1.18mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 31 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 60.0dB Ch2 Gain 2G 30.0dB
Ch1 Phase 1P 156.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
D2

15558/6/06
Copyright © 2006, TWI Ltd
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP +20dB Bal. Load LO ---
Graticule GR Rect.A
Lift off: 2.36mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 32 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 68.0dB Ch2 Gain 2G 30.0dB
Ch1 Phase 1P 156.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP +20dB Bal. Load LO ---
Graticule GR Rect.A
Lift off: 3.54mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 32 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 75.0dB Ch2 Gain 2G 30.0dB
Ch1 Phase 1P 156.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP +20dB Bal. Load LO ---
Graticule GR Rect.A
D3

15558/6/06
Copyright © 2006, TWI Ltd
Hot duplex plates Lift off: 2mm
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 33 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 60.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 310.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A

Settings: Probe 123P1D (2MHz)

Plate 316-01:
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 34 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 60.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 156.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A

D4

15558/6/06
Copyright © 2006, TWI Ltd
 Settings: Probe 102P1 (2MHz)

Plate: W1
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 35 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 46.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 144.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A
Plate: W1-1
Instrument used is Phasec 2200 Dual Frequency Instrument 33I012 S/W Version V10.19 04-Aug-98Standard 2
Dump 02 : 36 26 May '06
Probe PR Standard Mode MO Diff 1Ch
Display DI YT View VW X1
Ch1 Freq 1F 100kHz Ch2 Freq 2F 600kHz
Ch1 Gain 1G 43.0dB Ch2 Gain 2G 10.0dB
Ch1 Phase 1P 144.0° Ch2 Phase 2P 0.0°
Ch1 X:Y 1R 0.0dB Ch2 X:Y 2R 0.0dB
Sum Phase SP 0.0° Sum Gain SG 0.0dB
Hi-pass HP DC Lo-pass LP 2000 Hz
X-pos 1 1H 0 X-pos 2 2H 0
Y-pos 1 1V 0 Y-pos 2 2V 0
Alarm Shape AT Box Apply to AA Trace 1
Alarm Stretch AS 0.2s Alarm action AF Run Silent
Top TA Off Left LA
Right RA Bottom BA Off
Inner IA All Off Outer OA Off
Start SA 0.0° End EA Off
Analogue 1 Out A1 Off Analogue 2 Out A2 Off
Persist PE 1 sweep Sweep SD 5sec
Zoom ZM Normal Drive DR 0dB 2.0V
Inp. Gain IP 0dB Bal. Load LO ---
Graticule GR Rect.A

D5

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current analysis on Duplex plates
150

140

130

120

110

100

90
Amplitude

80

70

60
−♦− : 102P01
50

40 −■− : 801P15L
30
−*− : 800P01
−×− : 800P01JD1P
20

10

0 −▲− : 801P02
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Lift off (mm)

(a) : plate2, scan 15, detection of the crack 3

150

140

130

120

110

100

90
Amplitude

80

70

60
−♦− : 102P01
50

40
−■− : 801P15L
30 −*− : 800P01
−×− : 800P01JD1P
20

10

0 −▲− : 801P02
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Lift off (mm)

(b) : plate3, scan 10, detection of the crack 3

150

140

130

120

110

100

90
Amplitude

80

70

60

50 −♦− : 102P01
40

30
−■− : 801P15L
20 −*− : 800P01
10
−×− : 800P01JD1P
−▲− : 801P02
0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Lift off (mm)

(c) : plate4, scan 15, detection of the crack 7 and 8

Figure 1: Capability of Eddy current probes to work with lift off on duplex plate

D6

15558/6/06
Copyright © 2006, TWI Ltd
 150

140

130

120

110

100

90
Amplitude

80

70
−♦− : 102P01
−■− : 801P15L
60

50

40 −*− : 800P01
30
−×− : 800P01JD1P
−▲− : 801P02
20

10

0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5
Lift off (mm)

(d) : plate5, scan 24, detection of the crack 3 and 11

150

140

130

120

110

100

90
Amplitude

80

70

60

−♦− : 102P01
50

40

30 −■− : 801P15L
20
−*− : 800P01
−×− : 800P01JD1P
10

0
0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 −▲− : 801P02
Lift off (mm)

(e) : plate6, scan 19, detection of the crack 8, 6 and 9

Figure 1: Capability of Eddy current probes to work with lift off on duplex plate

D7

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current analysis on Stainless steel 316 plate
225

200

175

150

125
Amplitude

100
−♦− : 102P01
75 −■− : 801P15L
50
−*− : 800P01
−×− : 800P01JD1P
25
−▲− : 801P02
0
0 0,5 1 1,5 2 2,5 3 3,5 4
Lift off (mm)

(a) : plate 316-2, defect1

225

200

175

150

125
Amplitude

100
−♦− : 102P01
75
−■− : 801P15L
50 −*− : 800P01
−×− : 800P01JD1P
25
−▲− : 801P02
0
0 0,5 1 1,5 2 2,5 3 3,5 4
Lift off (mm)

(b) : plate 316-2, defect2

Figure 2 : Capability of Eddy current probes to work with lift off on stainless steel

D8

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current scans on Duplex plates with different lift off (Probe 800P01)
1). Lift off = 1.18mm (100kHz, phase: 115, gain: 48 dB)
Test block

Plate2

Scan 9: Crack 1 Scan 15: Crack 3 Scan 22: Crack 8

Plate3

Scan 10: Crack 3 Scan 14: Cracks 7 10 9 Scan 25: Crack 21

Plate4

Scan 9: Crack 3 Scan 15: Cracks 8 7 Scan 20: Crack 9

Plate5

Scan 10: Cracks 3 8 Scan 18: Cracks 10 3 11 Scan 24: Cracks 3 11

Plate6

Scan 11: Cracks 18 16 19 17 Scan 15: Cracks 10 12 Scan 18: Cracks 8 9 6

D9

15558/6/06
Copyright © 2006, TWI Ltd
2). Lift off = 2.36mm (100kHz, phase: 127, gain: 54 dB)
Test block

Plate2

Scan 9: Crack 1 Scan 15: Crack 3 Scan 22: Crack 8

Plate3

Scan 10: Crack 3 Scan 14: Cracks 7 10 9 Scan 25: Crack 21

Plate4

Scan 9: Crack 3 Scan 15: Cracks 8 7 Scan 20: Crack 9

Plate5

Scan 10: Cracks 3 8 Scan 18: Cracks 10 3 11 Scan 24: Cracks 3 11

Plate6

Scan 11: Cracks 18 16 19 17 Scan 15: Cracks 10 12 Scan 18: Cracks 8 9 6

D10

15558/6/06
Copyright © 2006, TWI Ltd
3). Lift off = 3.54mm (100kHz, phase: 127, gain: 63 dB)
Test block

Plate2

Scan 9: Crack 1 Scan 15: Crack 3 Scan 22: Crack 8

Plate3

Scan 10: Crack 3 Scan 14: Cracks 7 10 9 Scan 25: Crack 21

Plate4

Scan 9: Crack 3 Scan 15: Cracks 8 7 Scan 20: Crack 9

Plate5

Scan 10: Cracks 3 8 Scan 18: Cracks 10 3 11 Scan 24: Cracks 3 11

Plate6

Scan 11: Cracks 18 16 19 17 Scan 15: Cracks 10 12 Scan 18: Cracks 8 9 6

D11

15558/6/06
Copyright © 2006, TWI Ltd
4). Lift off = 4.72mm (100kHz, phase: 127, gain: 70 dB)
Test block

Plate2

Scan 9: Crack 1 Scan 15: Crack 3 Scan 22: Crack 8

Plate3

Scan 10: Crack 3 Scan 14: Cracks 7 10 9 Scan 25: Crack 21

Plate4

Scan 9: Crack 3 Scan 15: Cracks 8 7 Scan 20: Crack 9

Plate5

Scan 10: Cracks 3 8 Scan 18: Cracks 10 3 11 Scan 24: Cracks 3 11

Plate6

Scan 11: Cracks 18 16 19 17 Scan 15: Cracks 10 12 Scan 18: Cracks 8 9 6

D12

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current scans on Austenitic 316 plate 1 without lift off (Probe 123P1D)

(100kHz, phase: 156, gain: 60 dB)

Test block

Scan 9 Scan 10 Scan 11

Scan21 Scan22 Scan23

Scan24 Scan25 Scan26

Scan27 Scan28

D13

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current scans on Austenitic 316 plate 2 with different lift off
(Probe 800P01)

1). Lift off = 1.18mm (100kHz, phase: 156, gain: 60 dB)

Test block Plate 316-2

Scan1 (x=45mm) Scan2 (x=70mm)

2). Lift off = 2.36mm (100kHz, phase: 156, gain: 68 dB)

Test block Plate 316-2

Scan1 (x=45mm) Scan2 (x=70mm)

3). Lift off = 3.54mm (100kHz, phase: 156, gain: 75 dB)

Test block Plate 316-2

Scan1 (x=45mm) Scan2 (x=70mm)

D14

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current detection on Duplex welded plates (Probe 102P01)

1). Plate W1 (2Mhz, phase: 144, gain: 46dB)

Test block 3

13.3

40
Test block - 0.2mm deep flaw.
5
Plate Weld Defect 1 Defect 2

Defect 1 Defect 2

2). Plate W1-1 (2Mhz, phase: 144, gain: 43dB)

14

47
Test block
15

60

4 14
Weld Defect 1 Defect 2 Defect 3
Test block - 0.2mm deep flaw. Test block - 0.5mm deep flaw
Plate

Defect 1 Defect 2 Defect 3

D15

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current scans on cold and hot Duplex welded pipes

Cold pipe

Toe weld signal Crack area and toe weld

Hot pipe (100Hz, phase: 310 degrees, gain: 72 dB)

Signal of weld Signal of crack No crack

Crack No crack No crack

Crack Crack

D16

15558/6/06
Copyright © 2006, TWI Ltd
 Eddy current scans on hot duplex plates with the probe holder (Probe 800P01)

Lift off = 2mm (100kHz, phase: 310, gain: 61 dB), temperature controlled by thermocouples
type K (record with data logger)
Test block

Plate2

Scan11: Cracks 2,3 Scan 15: Crack3 Scan 17: Crack3

Scan 21: Crack8 Scan 23: Crack8

Plate3

Scan 6: Cracks 1,2 Scan 10: Crack3 Scan 14: Cracks 7, 5, 6, 8, 9, 10, 11

Scan 23: Crack21 Scan25: Crack21

D17

15558/6/06
Copyright © 2006, TWI Ltd
 APPENDIX E
ACFM Tests

15558/6/06
Copyright © 2006, TWI Ltd
 ACFM measurement on cold plate

Bx and Bz values 0 1 2 3
↓↓ ↓

Scan 1
x=160mm
Plate 2205-2

Scan 2
x=169mm

Scan 3
x=178mm

Scan 1
x= mm
Plate 2205-3

Scan 2
x=180mm

Scan 3
x=190mm

E1

15558/6/06
Copyright © 2006, TWI Ltd
 0 1 2 3
Bx and Bz values
↓↓ ↓

Scan 1
x*=175mm
Plate 2205-5

Scan 2
x=mm

Scan 3
x=195mm

Scan 1
x=166mm
Plate 2205-6

Scan 2
x=174mm

Scan 3
x=184mm

(*: distance from the side A of the plate)

E2

15558/6/06
Copyright © 2006, TWI Ltd
 ACFM Inspection Procedure

for AMIGO and QFM v3.x

Dr. Michael C. Smith

TSC Inspection Systems

6 Mill Square, Featherstone Road,
Wolverton Mill, Milton Keynes, MK12 5RB
UNITED KINGDOM

Document: TSC/MCS/1378
Date: 18 February 2002
Issue: 1.3 (12)
Revision: 12

© 2002. Technical Software Consultants Ltd.

 Document Revision Approval Sheet

6 Mill Square, Featherstone Road, Wolverton Mill,

Milton Keynes, MK12 5RB

KEYWORDS:

Controlled Copy: • Yes • No

TITLE:

ACFM Inspection Procedure

for AMIGO ACFM Instrument and QFM v3.x

Document: TSC/MCS/1378

MCS MCL DAT Diver use added.

1.3 ( 12 ) New probe types. 18 Feb 2002
Operator
certification.
Use of MPI.
MCL MCS MCL Modification to p15
1.2 ( 11 ) for v3.1 19 June 2000
MCL MCS MCL Created from
1.1 ( 10 ) 1118.doc 17 March 2000

Issue(Rev) Initiated by Checked by Approved by Date

Issued Document Revisions
Contents
INTRODUCTION 1
Overview.............................................................................................................1
Introduction to ACFM.........................................................................................1

DESCRIPTION OF HARDWARE 3
The ACFM Probes ..............................................................................................3
Amigo ACFM Instrument ...................................................................................4
Routine Maintenance and Handling.......................................................5
Manufacturers Contacts.........................................................................5

INSPECTION REQUIREMENTS 6
Introduction ........................................................................................................6
Personnel Required .............................................................................................6
Equipment Requirements ....................................................................................7

INSPECTION CONSIDERATIONS 8
Probe Deployment Considerations.......................................................................8
General Considerations - Welds.............................................................8
Circumferential Welds - No Ground Out Regions .................................8
Inspecting Ground Out Areas ................................................................9
Inspection During Grinding...................................................................9
Inspecting Small Tubulars .....................................................................9
Inspecting Flat Plate Welded Supports/Stiffeners.................................10
Rat-holes and plate ends ......................................................................11
Transverse Cracks ...............................................................................13
Material Features ..............................................................................................15
Surface Roughness ..............................................................................15
Coating Thickness...............................................................................15
Magnetic State.....................................................................................15
Surface Grinding or Work Hardening..................................................15
Seam Welds.........................................................................................16
Inspecting materials other than ferritic steel ........................................16

INSPECTION REPORTING 17
Reporting Requirements....................................................................................17
QFM3 ACFM Report Form...............................................................................18
Probe Operator Briefing: Check off list .............................................................19

INTERPRETATION OF ACFM SIGNALS 20

General Method ................................................................................................20
Crack Sizing .....................................................................................................23
Using MPI With ACFM ....................................................................................24
Length Sizing With ACFM .................................................................24
MPI Indications...................................................................................24
Use of MPI Lengths In ACFM Calculations.........................................26

1378.DOC Rev 12 Contents Page i

 OPERATIONAL PROCEDURE 27
System Setup.....................................................................................................27
Connections.........................................................................................27
Operations Check..............................................................................................27
Inspection Procedure .........................................................................................28
End Of Session Ops-Check And Backup Procedure...........................................30
Post-Session Ops-Check ......................................................................30
Backup Procedure................................................................................30
Exit QFM3 and check backup..............................................................30

APPENDIX A 31
Probe Specifications ..........................................................................................31
Standard Weld Probe...........................................................................31
Mini-Pencil Probe................................................................................32
Micro-Pencil Probe..............................................................................34
AMIGO Underwater Weld Probe.........................................................36
AMIGO Underwater Mini-Pencil Probe...............................................36
AMIGO Underwater Micro-Pencil Probe .............................................37
Underwater Pencil Probe .....................................................................38
Other Probe Types ...............................................................................39
Purpose of Probe Files .......................................................................................39

1378.DOC Rev 12 Contents Page ii

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

INTRODUCTION

Overview
This document describes the inspection procedure to be used with the Amigo ACFM instrument for manual
inspection with standard probes. These probes are designed to be deployed by hand, with data collected and
sent to a controlling PC. An example system is shown in Figure 1.
This procedure has been written for use with version 3.x of the QFM inspection software on a PC running
Windows 9x / ME / 2000 / XP or NT. To check the version of software in use, select the "About..." option in
the "Help" pull-down menu in QFM.

Figure 1 Amigo instrument using a laptop PC

Introduction to ACFM
The A.C. field measurement (ACFM) technique was developed from the A.C. potential drop (ACPD)
technique which has been used for crack sizing and crack growth monitoring. ACPD has been used
underwater even though electrical contact has to be maintained between the probe and the component being
inspected. The ACFM technique is simpler in operation as it depends on the measurement of the near-
surface magnetic fields rather than the surface electric fields, thus requiring no electrical contact.
Theoretical work carried out at the Wolfson NDE Centre in the Mechanical Engineering Department of
University College London determined the relationship linking these two fields. Thus existing models of
electric fields around cracks can be used to size cracks using magnetic field measurements. This non-
contacting sizing capability relies on the use of unidirectional input current in the region under inspection,
similar to that required for the ACPD technique. For the ACFM technique, the input current is induced into
the specimen thus making the system fully non-contacting.
In single probe ACFM operation the Crack Microgauge passes two signals to the ACFM crack detection and
sizing software (QFM). The first is the magnetic field strength measured in the direction parallel to the
crack edge (Bx) and the second is the magnetic field strength measured in a plane perpendicular to the

1378.DOC Rev 12 Page 1

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

surface of the metal (Bz). The software (QFM) then displays these signals in three forms; the Bx and Bz
traces separately against a timebase, a dual digital meter display, and a polar plot display in which one
component is plotted against the other. This latter form is known as a butterfly plot because of the
characteristic trace produced by a defect.
For more information, refer to TSC's introductory literature.
The ability to size surface breaking cracks without cleaning the weld region down to bare metal offers
significant potential benefits over existing techniques such as magnetic particle inspection, ACPD and eddy
currents. As well as allowing crack depth estimates to be made (for ferritic materials), the use of a
unidirectional input current provides further practical benefits. Firstly, the decay in strength of the input
field with probe height is relatively small so that variations in signal with probe lift-off are reduced.
Secondly, the current flow is arranged normal to a weld toe or other material discontinuity so that there is
no perturbation in current direction and hence no signal from the interface due to a change in material
property. A final benefit is that the technique requires no calibration for sizing. Techniques requiring
calibration rely on the measurement of signal strength on a standard notched sample. For weld inspection
the standard block is invariably of different material to that at the crack location leading to errors in
interpretation.

1378.DOC Rev 12 Page 2

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

DESCRIPTION OF HARDWARE

The ACFM Probes

A number of different probes are available for manual deployment in different geometries. Each probe
contains a sensor coil pair consisting of a Bx coil and a Bz coil wound concentrically. The sensors are
generally protected by a ceramic nose or disc to reduce wear.
The standard weld probe is the one normally supplied with the instrument. In general, the standard weld
probe should be used for detection and sizing whenever possible because it produces the most uniform input
field and is the most stable probe to deploy. Other probes are supplied depending on the geometry to be
inspected.
For tight angles, or other areas where the standard weld probe will not fit a number of miniature pencil
probes are available. However, it must be noted that miniature pencil probes are less able to differentiate
defects greater than about 5mm deep (compared to a limit of about 25mm for the standard weld probe).
Since the range of probes available is subject to change, TSC should be contacted for an up-to-date list of
probe types and their recommended applications.
Situations where the different probes should be used are highlighted in “INSPECTION
CONSIDERATIONS” on page 8.
Underwater probes are also available to allow the AMIGO to be used for certain inspections where the
instrument remains above the water line and the probe is deployed under the water. Note that only probes
designated as underwater may be used since topside probes are not generally waterproof.
Other special probes, including probes for use at elevated temperatures and other hazardous environments,
can be made to order if required by TSC.

1378.DOC Rev 12 Page 3

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Amigo ACFM Instrument

The layout of the Amigo instrument is shown in Figure 2.

Figure 2. Amigo Instrument Layout

The Amigo is designed to operate from its battery pack. The battery pack is recharged when necessary (on
or off the instrument) using a battery charger connected to the DC input socket (see Figure 3). The charger
runs on a mains supply (110 to 240 VAC, 50-60Hz).

Figure 3. Amigo with battery pack removed

Two sockets are available on the Amigo. One for connection of the inspection probe, the other for
connection of the RS232 communications cable to the PC.

1378.DOC Rev 12 Page 4

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Routine Maintenance and Handling

1. Note that the AMIGO is NOT waterproof and must never be submersed in water or other fluid. The
unit is rated to IP54 providing that all connectors or blanking plugs are in place.
2. Only probes specifically named as underwater probes may be used subsea. If the probe is to be used
in an environment where sharp marine growth could easily damage the probe cable, it is
recommended that the cable be protected using additional sheathing.
3. Extension cables may not be used with 50m probes.
4. Extension cables may not be used subsea.
5. When using subsea probes, clean probes with fresh water after each dive and dry before storing.
Ensure water does not enter the connectors.
6. The AMIGO instrument may be cleaned with a damp cloth.
7. Always keep connectors covered when not in use to avoid contamination.
8. Never put strain on the probe or connectors. If necessary, use "tie wraps" or other suitable means to
provide strain release in case of accidental tension being applied to the cables.
9. Refer to the AMIGO Maintenance Manual for further information.

** IF IN DOUBT, REFER TO MANUFACTURER **

Manufacturers Contacts
TSC Inspection Systems
6 Mill Square, Featherstone Road
Wolverton Mill South
Milton Keynes, MK12 5RB
UNITED KINGDOM

Tel: + 44 01908 220255

Fax: + 44 01908 220959
Email: [email protected]

1378.DOC Rev 12 Page 5

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

INSPECTION REQUIREMENTS

Introduction
This procedure has been developed to enable inspection of most geometries generally encountered in civil
structures, such as process plant, bridges, pipework etc. using standard manually deployed probes and QFM
software. For inspection using special-purpose probes or of any geometries not covered in this document,
please refer to TSC.
This procedure should only be used once the operator/supervisor is fully familiar with the following
documents:
1. QFM 3.x Software User Manual Level 1.
2. Amigo User Manual (TSC Doc. No. 1372).
Note: QFM User Manual Level 1 is intended for use by Level 1 (Inspectors) in order to collect, store and
analyse data. The QFM User Manual Level 2 is intended for use by Level 2 (Supervisors) allowing them to
change the parameters of the probe set-up and instrument control and to overwrite data files.

Personnel Required
ACFM Operator Experienced operator holding a CSWIP ACFM Level1 certificate or a Lloyds ACFM
certificate. Access to a CSWIP Level 2 is required if complex welds or interpretation
difficulties are encountered. If there is no Level 2 on site then reference may be made
to ACFM supervisory staff within TSC (see Manufacturers Contacts on page 5).
Probe Operator A technician who has been briefed by the ACFM operator on probe handling and
scanning procedures, and for whom the TSC Probe Operators check list has been
completed or
For subsea operation, a diver qualified to CSWIP 3.1U who has received training on
probe handling and scanning procedures and for whom the TSC Probe Operators check
list has been completed.
N.B. These two operators may be combined in a one-person operation if circumstances allow.
Whenever remote probe operators are used, continuous audio communication is required to enable the
ACFM operator to lead the inspection and for the probe operator to be able to report back on any local
factors influencing the inspection.
In addition, if an inspection is carried out with divers then a diver helmet-mounted camera system must be
used to enable the topside operator to supervise the subsea inspection site.
Operators holding certification from other recognised Qualification / Certification schemes than the above
may be considered suitably qualified, however they should contact the qualification body responsible for the
inspection site for official approval.

1378.DOC Rev 12 Page 6

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Equipment Requirements

• Amigo Portable Crack Microgauge

• RS232 Serial Communications cable
• ACFM Probes - Types selected according to geometry and use
• ACFM Probe Extension cable (optional). Not for use with subsea probes.
• RS232 Extension Cable (optional)
• Spare Amigo battery pack and/or battery charger (optional)
• Multi-Point Block and Cable with 110V or 240V plug (to be nominated by operator)
• Canvas Carry Case
• Transit Case
• Marker Pencils / chalk (Box) or magnetic arrows
• Ruler / Measuring Tape
• Operations Check Block
• Test Sample of the same material type as that to be inspected if not ferritic steel.
• Laptop portable PC with 3.5" Floppy Disk Drive (A:) and RS232 comms port on 9-pin 'D'
connector, running Windows (9x, ME, 2000, XP or NT) and QFM Software v3.x installed on hard
disk.
• Spare Computer Battery and Charger
• Probe File Disk for all ACFM Probes*
• Box of Formatted Floppy Disks
• Operations Manual For Amigo
• QFM Vn3.x Software Manuals Level 1 and 2
• TSC ACFM Report Sheets
• Wireless 2-Way Communications Set (optional for 2 person operation)

*N.B. The TSC issued probe files delivered with the probes are set up for use on ferritic steels. If they are to
be used on another material they may need to be modified to ensure that the signal levels obtained are
correctly displayed on the screen. The procedure for doing this is described in the Level 2 Software User
Manual.

1378.DOC Rev 12 Page 7

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

INSPECTION CONSIDERATIONS

Probe Deployment Considerations

General Considerations - Welds

A number of probe types are available for use with the Amigo, including a standard weld probe and several
variations of mini and micro-nosed pencil probes. In general, the standard weld probe should be used for
detection and sizing whenever possible because it produces the most uniform input field. However, the nature
of many inspection geometries means that the other probes must often be used as directed below.
The following notes on probe deployment describe how to inspect welded components for fatigue cracks
where it is assumed that defects will closely follow the weld line. The technique relies on recognition of the
signal from a probe scan along the length of a defect, so an ACFM probe is always scanned along a line
parallel to the weld. For this reason, defects that lie at an angle of more than 45o to the weld may not be
detected with this procedure. If inspection for transverse defects is required, refer to the procedure in
“Transverse Cracks” on page 13.
The recommended scanning speed is about 50mm per second. Slower speeds may be used, especially subsea,
and the ACFM trace speed adjusted accordingly. The standard probe scans a width of approximately 15mm.
Scans should always be made along both weld toes and, if wider than 15mm, the weld cap should also be
covered by making a number of passes sufficient to cover the weld cap width taking into account the coverage
of the probe.
If a defect indication is found during a weld toe scan, repeat the scan in the first weld bead to identify whether
the crack indication is due to inter-bead cracking. If the indication during the bead scan is larger than that
observed during the weld toe scan, the defect must be in a weld bead. Repeat in the adjacent beads until the
defect indication is greatest. The bead giving the greatest signal is the one containing the defect. Note that
inter-bead cracking often jumps between beads, which will result in a sudden drop in the signal from one
bead coinciding with a sudden rise in the adjacent bead.
In some geometries, especially tubular intersections, long defects that start growing along a weld toe can
deviate into the parent plate. This will show itself as a sudden drop in Bx with no corresponding Bz peak or
trough. In this case, scans should be made on the parent plate, along the expected defect line, to find the
defect's end.
Where the crack depth calculated is greater than the plate thickness this indicates that cracking may be
completely through the weld and so the probe operator should be asked to look for further evidence of this.
Such evidence may be visual crack opening or, if the back face is accessible (e.g. on flat plates), crack-like
indications there.

Circumferential Welds - No Ground Out Regions

Standard weld probes should be used for all circumferential welds where access allows. If the chord-to-brace
angle is too small to allow access of the standard probe, a probe for use in tight access areas should be used. A
pencil probe should only be used for inspecting ground out regions or other geometries where a weld probe
cannot gain access. Small diameter welds are normally inspected with a pencil probe.
Where the length of weld to be inspected is greater than 400mm the weld should be marked up into smaller
lengths which overlap. (Longer scans tend to be more difficult for the probe operator to perform, therefore
increasing the chance of probe lift off etc.). The measurements from a datum to the four cardinal clock
positions should be recorded. The location of any features that affect probe movement or probe signal should

1378.DOC Rev 12 Page 8

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

be reported (e.g. the position of a seam weld that joins the weld under inspection, or the presence of weld
spatter).
When a crack is located it should be re-scanned in more detail by scanning the defect area plus 30mm before
and after each end at a slower scan rate. The crack tips should be located and marked and surface breaking
crack length reported to the PC operator. The circumferential distance from a datum (e.g. 12 o'clock
position) to one end of the crack should also be noted.
If the weld is to be ground out to remove the defect the estimated depth using ACFM should be the depth to
which the first grind is made. Before grinding takes place the defect plus 30mm either side should be
scanned with a Pencil probe for later comparison. After grinding the grind should be re-inspected with the
same Pencil probe to ensure the defect has been removed (the pencil probe type selected for the pre-grind scan
should therefore be chosen to fit into the subsequent ground area).

Inspecting Ground Out Areas

Welds that have been ground to remove defects require the use of a Pencil probe. A pencil probe is necessarily
easier to tilt or rock than a weld probe and so extra care must be taken during a scan to ensure that this does
not happen. The first scan with the probe should be with the probe axis perpendicular to the root of the grind.
The weld should be scanned such that at least 30mm either end of the ground out region is scanned. If a
defect signal is found two further scans should be performed to determine whereabouts around the grind the
defect is located:
a) Probe axis at +30o angle from the normal to the root of the grind.
b) Probe axis at -30o angle from the normal to the root of the grind.
The maximum signal perturbation should be used to determine crack depth and should also indicate the
location of the defect in the ground out channel.
The ground out region of the weld should be marked at suitable intervals (e.g. 20mm) which should be called
to the PC operator as the probe scan passes each mark. A measurement of length from a suitable datum (e.g.
12 o'clock position) to the weld grind marks should be made and also reported to the PC operator.
Where the geometry of the grind start or finish is sharp this may obscure small defects. The PC operator will
advise whether further grinding is required to smooth off the edges of the grind to give good probe movement
and signals at the grind start.

Inspection During Grinding

If a detected defect is to be removed by grinding, a scan of the defect using a pencil probe must be made prior
to any grinding. This provides a baseline for comparison, removing the effects of variations in sensitivity
between different probes. The whole of the grinding process should be monitored with the same probe.
Between grinding passes, the defect should be rescanned. Particular note should be made of the position of the
defect in the grind profile, to ensure that the grind follows the defect even if the defect curves away from the
expected direction.
The topside operator should note changes in the ACFM signal between grinding passes, such as changes in
length, or splitting into two or more separate cracks. An apparent increase in length or depth may indicate a
sub-surface or branching crack.
If MPI is being used in addition to ACFM, care must be taken when comparing the results for sizing
purposes. Refer to the section "Length Sizing With ACFM" on page 24 for the procedure to be followed in
this case.

Inspecting Small Tubulars

For small tubular joints the use of a pencil probe is recommended. To avoid excessive probe rock and to keep
the probe nose in the same position relative to the weld, a probe shoe can be used, moulded from such
material as plastic padding, which is shaped to fit the weld contour.

1378.DOC Rev 12 Page 9

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Inspecting Flat Plate Welded Supports/Stiffeners

Full geometry detail is required by the PC operator before scanning can commence. As with tubular
components, the standard weld probe should always be used where geometry allows and other probe types
only used where necessary.
The use of a uniform input field, which allows defect sizing, means that the standard weld probe picks up
strong signals from sharp geometry changes such as plate edges, corners and rat-holes. This edge effect
occurs when a probe is used within a certain distance of the edge, the distance being roughly equal to the
width of the main part of the probe body. Thus a standard weld probe will pick-up edge signals when the
mid-point of the probe is within about 50mm of the edge. A pencil probe also experiences edge effects, but to
a much lesser degree. Mini/micro pencil probes can be used up to about 10mm from an edge.
The presence of an edge does not preclude the use of standard probes but signal interpretation is not so easy.
The effect of an edge is to superimpose a general slope in probe readings that will increase in gradient as the
plate edge is approached. For detection, this has the effect of separating the signal from either end of a defect
on the butterfly plot so that the loop formed is not closed. Where cracks are located within 50mm of a plate
edge the depth of the defect should be calculated using an estimate of what the background Bx would have
been at the centre of the defect, as shown in Figure 4.

Bx background

Bx minimum

Figure 4. Bx background and minimum values to be used near a plate edge.

If there is not too much curvature in the signal, this background Bx is best estimated as an average of the Bx
values either side of the defect.
Instead of clock position marking, the weld to be inspected needs to be marked off in suitable linear intervals
(maximum 100mm).

1378.DOC Rev 12 Page 10

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Rat-holes and plate ends

On rat-hole specimens with full welds (as shown in Figure 5), the mini pencil probe should be used if the
weld lengths involved are all short.

Top view Side view

Figure 5. Views of rat-hole with complete welds (welds shown hatched)

The probe should be first used to detect longitudinal defects in a series of 6 scans as indicated in Figure 6.
The scan directions are such that the probe is always moving in the 'C' direction so that a defect will be
indicated by a clockwise going loop in the butterfly plot. Note that when the probe leaves or approaches a
weld, a change in the Bx level is seen. There is no significant associated change in Bz however, so the signal
should not be confused with a defect signal.

3 4

6
5

1 2
I.D.

Figure 6. Detection scans 1 to 6 for rat-hole specimen with complete welds

With any scan, a defect lying wholly along the scan path will give a complete loop. However a defect leaving
the scan path will give a lop-sided loop because one or both of the defect ends may not be covered. A defect
leaving the scan path will show a sudden return in Bx to the background level but no accompanying Bz peak
or trough.
Scan 1, for instance, starts on the parent plate about 25mm from the corner of the weld. A defect running into
the parent plate from the weld toe (marked as A in Figure 7) will thus give a complete loop as shown, while a
defect that follows the weld toe round the corner (marked as B) will give a lop-sided loop.

1378.DOC Rev 12 Page 11

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Traces for scan along path a

Defect A Defect B

Bx

B
Bz
A

a
Butterfly

Figure 7. Comparison of signals from a defect that follows weld to one that leaves weld

Similarly, the end of scan 1 in Figure 6 follows the weld toe round to the mid-point of the rat hole, so a defect
following the weld toe will give a complete loop, while a defect continuing into the parent plate will give a
lop sided loop. In either case, a later scan should be taken along the suspected line of the defect to confirm
the defect and to size it.
Scan 5 (or 6) will give a complete loop for a defect running into the parent plate. For a defect running around
both corners, only a dip in Bx will be seen coinciding with the extent of the weld. In this case, the defect ends
will be seen in scans 1 and 3 (or 2 and 4). Again a scan around the weld corner must be made later to confirm
the defect. When no defect is present, scans 5 and 6 show a low peak in Bx. This produces a dent in the
trough produced by a defect as shown in Figure 8.

Traces from defect shown

Bx

Butterfly

Figure 8. Signal from a short defect at a free plate end

Finally, scans should be made as shown in Figure 9 to detect defects growing into the parent plate from the
rat-hole welds. On specimens with large rat-holes, the scans can be combined into two scans (i.e. join 7 and 8
together, and 9 and 10). For smaller rat-holes, the right-angled mini pencil probe should be used. The scans
should be continued into the parent plate until the traces become flat - i.e. about 25mm if no defect is present.
Four scans will be necessary for specimens with very small rat-holes where there is insufficient room for the
probe to pass through, but scans 7 and 8 (and 9 and 10) should be combined on one page by pausing the scan
between lifting the probe from the surface at the end of scan 7, and putting the probe down at the start of scan
8.

1378.DOC Rev 12 Page 12

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

8 9

7 10

I.D.

Figure 9. Detection scans 7 to 10 for rat-hole specimen with complete welds

Having completed all the detection scans, positions and lengths should be measured with the same probe, and
then defects should be sized with the standard weld probe if possible. Defects inside the rat-hole can be sized
with the mini pencil probe. The procedure to be followed is therefore:
1. Use the standard weld probe or mini pencil probe to do 10 scans as described above.
2. If defects suspected to run away from scan line, carry out additional detection scans as described
above.
3. Measure the position and length of all defects found in steps 1 and 2.
4. Change to the standard probe, if necessary, and scan over all defects found and measure their depths.
For rat-holes and plate ends with incomplete welds, a similar procedure should be followed, but no scans
should be made where no weld exists (e.g. scans 5 - 10 above).

Transverse Cracks
If the geometry or conditions are such that it is suspected that the weld may contain transverse defects, the
following probe deployment procedure should be carried out for detection. In order to obtain a signal from a
transverse defect, the input field must run perpendicular to the defect, so to detect a transverse defect the
ACFM probe must be turned around by 90o as shown in Figure 10. The standard weld probe should be used
for flat or cylindrical butt welds; for T-butt welds where deployment of this probe is not possible, a pencil
probe should be used.
The normal probe scan path to produce a butterfly plot loop would then be across the weld cap, however for
detection this would have to be repeated every 10mm or so along the weld resulting in a slow detection
process. Instead of this the probe should be scanned along the centre of the weld cap as shown in Figure 10.

butt weld T butt

Figure 10 . Probe orientations and scan directions for initial detection of transverse defects across
weld caps

1378.DOC Rev 12 Page 13

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Since the scans are not along the expected line of a defect, a transverse defect will not produce a loop in the
butterfly plot. Instead, the possible presence of a defect is indicated by a dip in the Bx signal. The Bz signal
may show a peak or a dip if the scan passes nearer to one end of the defect than the other, but will show
nothing if the scan passes over the middle of the defect. The dip in Bx will be about 20mm across - this width
is determined mainly by the probe input field geometry rather than the defect size. The depth of the dip is
determined by the defect depth.
Since probe rock and lift-off changes will inevitably be more severe when the probe is scanned along the weld
cap, the presence of a dip in Bx is not enough on its own to indicate the presence of a defect. In order to
confirm a defect-like indication, a scan should be made across the weld cap along the line through the dip
found in Bx (i.e. along the line of the suspected defect) as shown in Figure 11.

butt weld

defect

Figure 11. Probe orientation and scan direction for confirming detection of transverse defects across
weld caps

The probe should still be oriented as before, and the scan should start and finish at least 25mm from the weld
toes to ensure that both defect ends will be covered. In this way, a defect will produce a loop on the butterfly
plot as for conventional weld toe defects. Note, however, that since any material property change associated
with the weld now runs across the scan direction, signals may be seen in both Bx and Bz due to the weld (this
is similar to scanning across a seam weld in conventional deployment). Since these weld signals may produce
confusing loops in the butterfly plot, a scan must also be made across the weld cap away from the suspected
defect for comparison.

1378.DOC Rev 12 Page 14

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Material Features

Surface Roughness
The surface should be cleaned to allow smooth probe travel which requires removal of loose flaking corrosion
or marine fouling. Hand cleaning with a scraper is normally sufficient. The system operator shall confirm the
surface condition is acceptable prior to carrying out the inspection.

Coating Thickness
With the ACFM technique, it is possible to detect defects through 5mm or more of non-conducting coating.
The signal strength obtained is obviously smaller than for an inspection with no coating but sizing estimates
can still be obtained by entering an estimate of the coating thickness in the QFM sizing procedure using the
box labelled "Extra lift off". The look up tables presently implemented do not support extra lift-offs greater
than 3 to 5mm (dependent on probe type) - refer to TSC for sizing through thicker coatings. Note that,
because of the use of a unidirectional input field, the ACFM signals are relatively insensitive to lift-off and so
the thickness need only be estimated to the nearest millimetre. For the same reason, the ACFM technique is
relatively insensitive to changes in coating thickness (or changes in lift off) and in any case, only the Bx
signal is affected so no confusion with a possible defect signal can occur. It should be borne in mind that the
reduction in Bx and especially Bz amplitudes due to coatings will result in smaller butterfly loops than
otherwise expected for a given defect size. To compensate for this, a new probe file should be made through a
non-conducting layer of the same thickness on a standard ops-check sample.

Magnetic State
It must be ensured that the surface being inspected is in an unmagnetised state. Therefore the procedure
followed with any previous magnetic technique deployed (particularly MPI inspection) must include
demagnetisation of the surface. This is because areas of remnant magnetisation, particularly where the leg of
an MPI yoke was sited, can produce loops in the butterfly plot which may be confused with a defect signal.
If magnetisation of the structure is unavoidable (e.g. d.c. magnets used to attach an ROV), the magnetised
area must be at least 500 mm away from the weld to be inspected.

Surface Grinding or Work Hardening

It has been noted in a number of instances, especially on Duplex, that areas of surface grinding or work
hardening etc. which produce localised changes in material properties can result in strong signals in both Bx
and Bz which may be confused with a defect signal. Areas of grinding should be reported by the probe
operator at the time of scanning and recorded by the ACFM operator.
If a defect signal is suspected in a region of grinding or possible work hardening, further scans should be
taken parallel to but away from the weld toe. The signal from a defect (especially Bz) will drop off quickly
away from the defect, so that a scan away from the weld toe will be much flatter. The rate at which the signal
drops off will depend on the length and depth of the crack, but a significant difference in amplitude should be
seen 20mm away if the signal is due to a defect. If there is no significant change in signal amplitude 20mm
away from the weld toe, the signal is likely to be due to the effects of the grinding. A further scan 40mm
away should be made to confirm the trend. To assist later signal interpretation, marker lines should be added
on the display at the start and end of the ground area.
If an edge effect probe is available, and geometry allows its deployment, a scan with this probe will produce a
much flatter signal from grinding marks than from a defect.

1378.DOC Rev 12 Page 15

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Seam Welds
Seam welds running across the line of scanning also produce strong signals in Bx and Bz which can
sometimes be confused with a defect signal. As for grinding marks above, if a defect is suspected at or near a
seam weld, further scans should be made away from but parallel to the suspected defect line to confirm the
defect. The signal from the seam weld will persist at roughly the same amplitude, whereas a defect signal
will exhibit a much reduced amplitude. Again, an edge effect probe, if available, can be used to emphasise
any defect signal.

Inspecting materials other than ferritic steel

The probe file associated with a given probe contains information used to produce accurate sizing estimates,
and scaling information to match the screen display limits to the signal levels produced by the probe on a
reasonable size defect. By default, these scales are set so that a 50mm long by 5mm deep notch in ferritic
steel produces a loop on the butterfly plot that is roughly centred on the screen, fills about 50% of the height
of the plot (i.e. Bx amplitude) and more than fills the width (Bz amplitude).
If a probe is to be used on other materials, such as Duplex, aluminium, stainless steel, etc. the scaling may not
be correct so that the display may be more or less sensitive resulting in the possibility of getting more spurious
indications or missing small defects respectively. Also, the depth sizing will not be accurate. In order to
avoid this a new probe file must be used for the material in question. In the first instance contact TSC to see
if a relevant probe file is available. If not, refer to the QFM v3.x Level 2 Software Manual.

Duplex
Duplex is known to be more susceptible than other materials to spurious signals caused by material
differences arising from grinding, heat treatment etc. In some cases it is possible for these signals to produce
butterfly plot loops similar to those produced by a defect. For this reason defect signals should be confirmed
by taking extra probe scans adjacent and parallel to the suspected defect as described in “Surface Grinding or
Work Hardening” on page 15.

1378.DOC Rev 12 Page 16

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

INSPECTION REPORTING

Reporting Requirements
ACFM techniques produce floppy disks full of inspection data. These will only be useful if backed up with
data sheets filled in correctly. This section deals with the information required and produced on these data
sheets.
The data report sheets generated by ACFM inspections will be specifically designed with the ACFM system
and current inspection requirements in mind. An ACFM data sheet should contain all the following:

General Information
Date
Operators Name
Probe Operator
Component ID number
File Number
Scanning Data
Filename
Page Number
Position on Riser
Probe Number
Probe Direction
Clock or Tape Positions
Inspection Summary
Detailed Record of Indications/Anomalies
Filename
Page Number
Position on Riser
Start of Defect (Tape reference)
End of Defect (Tape reference)
Length of Defect (in millimetres)
Remarks
Diagram/Drawing of component under inspection.

A poorly filled in report sheet reflects badly on the operator, and the equipment, and will render off-line
interpretation difficult if not impossible.

1378.DOC Rev 12 Page 17

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

QFM3 ACFM Report Form

Location: Sketch of geometry:

Date:

Time:

Operator: Probe Op.:

Component ID:

Summary of indications:

Filename:
Probe number/configuration: Probe file:

Distance
Direction of Circumf. Page Inspection report / comments
From
Datum travel A/C Position

1378.DOC Rev 12 Page 18

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Probe Operator Briefing: Check off list

Operators name
Date
Probe operators name

Tick boxes accordingly

Probe YES NO
1) Shown various probe selections and reasons for use
2) Shown probe connections/connectors to instrument
3) Shown probe identification markings and sensor centre-line
4) Instructed on leading edge of probe direction ie A or C
5) Shown how to hold the probe

Scanning techniques YES NO

6)
1) Instructed on marking
Shown various up component
probe selections under inspection
and reasons for use
7) Advised on scan starting procedure
8) Advised on scanning rate and indicating circumf. posn.
9) Instructed on overlapping procedure

Visual Inspection YES NO

10) Made aware
1) Shown variousofprobe
visualselections
defects that
andmight influence
reasons scans
for use
11) Probe fit for tight geometries

Marking and Sizing Defects YES NO

12) Instructed
1) Shown on how
various to selections
probe size defects forreasons
and length for use

1378.DOC Rev 12 Page 19

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

INTERPRETATION OF ACFM
SIGNALS

General Method
In general, a defect will product a characteristic signal on Bx and Bz signals and these combine to give the
"butterfly". The rule adopted is that the signal represents a defect when
1) Bz responds by a peak and trough, and
2) Bx responds by a "dip" from the mean value.
In the majority of situations for, isolated and relatively short cracks, this will give a butterfly plot which
moves to right or left (according to probe direction), then downward then returns to the original position to
close the loop. Figure 12 shows the relationship between Bx, Bz and the butterfly trace where "A"
represents the first Bz peak, "B" represents the deepest point in Bx, and "C" represents the Bz trough.

Chart recorder plots

Bx

B
Time

A
Bz

Bx
C
A

B
Bz
Butterfly plot

Figure 12.Schematic showing relationship between butterfly and chart recorder plots

Lift off, permeability changes, probe rocking etc. can cause responses from Bx and/or Bz. Lift-off, for
example, causes the Bx signal to dip or peak (depending on material type) with little response from the Bz
signal. This would produce a closed loop confined to the vertical axis rather than the open loop produced by

1378.DOC Rev 12 Page 20

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

a crack. A seam weld, on the other hand, usually causes a peak in Bx combined with Bz signals that result
in an open loop moving upwards from the starting point. Hence, for the signal to represent a crack, the
butterfly loop must move downwards. This rule is intended to eliminate false calls due to momentary lift
off, perhaps as the probe runs over local weld imperfections. It therefore is intended for small signals.
There are exceptions to the general rules and these mainly apply to long defects (> 50mm).
For long cracks the Bz peak and trough may be some way apart. This means that at the centre of the crack
there is no Bz signal and we rely on Bx (i.e. the dip in Bx) to make the butterfly go down. If there is a
general drift on the Bx signal, i.e. if the top signal is not perfectly level, this may confuse the butterfly by
acting against the crack signal. This conflicting Bx signal means the butterfly rules no longer can be relied
on and it is necessary to use the Bx and Bz plots to look for tell-tale signals where there is a response from
Bz followed by a downward deviation, from the trend, on Bx. This is particularly important for tight
angles where the Bx signal trend rises toward the tight angle due to global geometry effects.
In theory it is possible for a weld to be cracked around the full circumference, thus resulting in no crack
ends. If the crack has a uniform depth, the Bx signals would be lower than expected, but in all other aspects
the signals could be similar to an uncracked connection. In this situation the presence of a signal centred
much lower on the butterfly plot than for other connections should be an indication that a full circumference
defect is present. This can be further investigated by comparing the Bx signals as the probe is moved from
parent plate to the weld toe area and then repeating the exercise on the other weld toe.
In practice, cracks do not tend to grow this way in tubular connections. Experience has shown that full
circumferential cracking is normally associated with significant variations in crack depth around the
connection combined with crack branching. In this situation the crack branching provides Bz signals
resulting in butterfly loops in the normal way. Thus in practice detection of full circumferential cracks on
node connections is similar to normal cracks in that butterfly loops will normally be present together with
significant depth variation represented by dips in Bx and downward movement of the butterfly trace.
Crack depth measurement is more complicated if full circumferential cracking is suspected because the
"length" and background Bx readings will not be easily established. Refer to TSC if assistance is needed.

1378.DOC Rev 12 Page 21

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Yes
START Downward Butterfly ? CRACK

No

No Yes
Crack Unlikely
Any Significant Possible Crack
Loops ?

Any Dips in Bx
No Below the Trend ?

Yes
No
NO CRACK Any Associated peaks If it is possible that the probe was not
and troughs on Bz? following the weld, rescan.
Otherwise, NO CRACK
Yes

Probably a long CRACK.

Scan a large section and look for crack-like
trends, especially on Bx, to confirm.

Figure 13. Decision Tree for Identification of a crack

At all times look at the Bx and Bz traces. If Bx has a dip then suspect a crack. If the butterfly makes any
significant loops, look at a broad area either side of the signal. This is particularly important if the butterfly
is moving up or down the screen. A butterfly moving up or down the screen with any sort of looping is
likely to be a long crack.

1378.DOC Rev 12 Page 22

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Crack Sizing
Crack sizing can be carried out either on-line (immediately after the necessary data has been collected) or
off-line (by recalling stored data). During the detection stage, scans of Bx and Bz are recorded for all
defects found. A first estimate of length is also obtained from a manual measurement of the extent of the Bz
signal. This data is then used to obtain sizing estimates using the QFM program as outlined below. For
more details, refer to the Level 1 Software User Manual.
1. Select a section of the Bx timebase plot to one side of the defect signal that is representative of the
background level. This should simultaneously be near the centre of the area filled by the general
off-crack background noise in the butterfly plot. This value is chosen as the Bx Background Level.
N.B. If the background level is significantly different either side of the defect, the average of the
two values should be used. This is normally necessary for defects in regions of changing geometry
or near to a plate edge (see Inspecting Flat Plate Welded Supports/Stiffeners).
2. Select the minimum of the Bx timebase trace from the centre of the defect signal (or the bottom of
the butterfly loop). This value is chosen as the Bx Minimum Level.
3. To calculate defect length and depth, the QFM software requires three parameters - the two Bx
values selected above, and the length estimate obtained during the detection stage (the length in
mm between the Bz peak and trough indications at the inspection site).
Where the crack depth calculated is greater than the plate thickness this indicates that cracking may be
completely through the plate and so the probe operator should be asked to look for further evidence of this.
Such evidence may be visual crack opening or, if the back face is accessible (e.g. on flat plates), crack-like
indications on a scan made there. Alternatively, because the currents follow the crack faces exactly in
ferritic steel, depths apparently greater than the wall thickness can arise from crack branching into the
parent plate, or from a highly curved crack path.

1378.DOC Rev 12 Page 23

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Using MPI With ACFM

When defects are being removed by grinding, it is often the case that MPI is used in conjunction with
ACFM to monitor defect removal. In these cases it is tempting to save inspection time by using MPI lengths
instead of Bz lengths when sizing defects. This will lead to errors and is not a recommended practice. This
section considers the implication and suggests how, under certain conditions, data from MPI inspection can
be used in the absence of any additional information.
When conducting MPI inspections an approved MPI inspection procedure should be used.
De-magnetise the area following MPI inspection (see Magnetic State on page 15.

Length Sizing With ACFM

The method of length sizing involves determining the location of the Bz peak and trough signals on the
component. The distance between peak and trough is referred to as the Bz Length. The ACFM modelling
shows that the Bz peak/trough actually occur just inside the physical ends of the crack. This is confirmed in
practice when the crack ends are marked and compared visually. Thus the Bz lengths are always less than
the actual crack length. This is taken into account in the sizing algorithms so that the ACFM calculated
length is greater than the Bz length.
Because ACFM also considers the depth of the crack, ACFM will "concentrate" on those parts of a crack
with significant depth.
If two cracks are joined at the surface by a shallow surface crack, ACFM will focus on the two deep parts,
not the whole, and ACFM sizing procedures take this into account.

MPI Indications
MPI will simply indicate the length of a surface flaw and give no indication on depth. MPI lengths should
not be used as a substitute for Bz lengths because:
1. For any crack MPI length > Bz Length. (Figure 12)
2. The crack may be made up of more than one deep section. MPI will treat it as one crack and
therefore misrepresent the length. (Figure 13)
3. The crack may have long "tails", again misrepresenting the length of the significant crack. (Figure
14)

1378.DOC Rev 12 Page 24

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Bz Length

MPI Length

Figure 12. Comparison of Bz peak/trough with actual crack ends

Bz 1 Bz 2

MPI Length

Figure 13. ACFM Bz signal compared to MPI indication for two adjacent cracks

1378.DOC Rev 12 Page 25

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Bz

MPI Length

Figure 14. ACFM Bz signal compared to MPI indication for crack with shallow ends.

Use of MPI Lengths In ACFM Calculations

One example where it may be preferable to use MPI lengths instead of Bz lengths is a situation where
ACFM is being used during a grinding process and it is found that there is a start/stop Bz signal at the grind
profile ends which are close to the defect ends. Hence there may be some doubt as to whether the crack Bz
locations are being masked by the grind profile end signals. As an alternative, MPI lengths may be used in
the crack depth sizing. As described earlier, the use of MPI length as a substitute for Bz length is incorrect.
However there is a work around if, and only if, certain criteria are met. The answer will still be
approximate but may be acceptable. The criteria are:
1. Before grinding, carry out an ACFM inspection and use Bz to calculate the crack length.
If:
i) The calculated length agrees with the MPI length, and
ii) The Bx signal indicates a simple crack
continue to step 2. If either of these conditions are not satisfied, do not attempt to use MPI
information with ACFM.
2. Estimate by how much the Bz length is less than MPI. You should find that the Bz length is
between 80% and 90% of MPI length.
3. Carry out the grinding.
4. Re-inspect with ACFM. If the crack remains a single crack (i.e. no additional features appear on
Bx) take the MPI length after grinding and apply the factor in (2) above. Then compare the
estimated ACFM length with the factored MPI length. If they agree, the factored MPI length
should give a reasonable depth estimate when used in the ACFM sizing procedure. If the two
estimates do not agree, it is possible that a situation such as shown in Figure 12 has arisen and the
MPI length should not be used to depth size the defect.
5. The above process can be repeated providing the crack remains as a single defect. Should it split
into two, the process will not work and any depth calculation will be in error.

1378.DOC Rev 12 Page 26

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

OPERATIONAL PROCEDURE

System Setup

Connections

Amigo Instrument
The Amigo operates on its own battery pack, although this can be powered from 110/240V AC (50/60Hz)
mains through a charger if necessary. The instrument communicates with the PC through an RS232 cable
which must be connected to the 8 way Lemo socket on the instrument panel. The probe to be used is
connected to the adjacent 18-way Lemo socket.

Computer
The computer operates on its own internal battery or on 110/240V AC (50/60Hz) through a mains pack /
battery charger. Refer to the computer user manual for details on how to connect the mains pack. The
computer communicates to the Amigo via the RS 232 cable supplied. The RS232 port is usually a 9-way D-
plug on the rear of the computer. Refer to the computer user manual for more details.
NB. To maximise Ni-Cad battery life always run PC on batteries until flat, then change battery and charge up
original battery outside the PC using the battery charger. Do not 'top up' the charge on a partly used battery.

Connection Procedure
1. Connect the computer to the Amigo using the RS232 cable (8-way Lemo to 9-way 'D'-type).
2. If battery power is low and no spare batteries are available, connect the Amigo and/or computer to
mains supply.
3. Switch on the Amigo by pressing down the power button on the front panel for two seconds. Check
that the red led starts flashing. Switch on the computer and start QFM software (as described in
Software Level 1 User Manual).

Operations Check
At the start and end of an inspection session an operations check must be performed using the equipment.
This is to ensure that the equipment is functioning correctly and to familiarise the operator with the relative
levels of noise and defect signal. The following procedure assumes familiarisation with the QFM software.
For details, refer to the appropriate Software User Manuals.

1. Select Probes for next series of scans to be carried out (see “Probe Deployment Considerations” on
page 8). If not already done, connect instrument as described in System Setup - Connection
Procedure.
2. Create a directory to store the data for the new inspection session.
3. Start the QFMv3 software, and check that Amigo is initialised correctly.
4. Connect the first probe to the instrument and select the appropriate configuration from those
available for that probe.

1378.DOC Rev 12 Page 27

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

5. Create a new data file for the operations check data.

• The format for the file name should be
OPpppnn.WDF where
O (Oh)P stands for OPs-check
pppp is the probe serial number (e.g. 1734)
nn is the number of the operational check of that day, i.e. first ops check in the day would
have n = 01.
6. Place probe on the test block and start data collection.
7. Scan along the test block slot at approximately 50mm/sec. After passing the slot stop the scan.
Signals such as that shown in Figure 14 should be displayed. Note that small differences in material
or gain used between creating the probe file and doing the ops check may mean that the display is
just off screen, in which case the signal should be centred on the screen.

Figure 14.Typical plots for ops check on a single defect

If the display is well off the screen, check that the correct configuration was chosen for the material
of the ops check plate, and that the serial number of the probe displayed by the software matches that
marked on the probe. If no configuration matching the probe and/or the ops check plate material is
available, a new configuration file must be set up, as described in the Software Level 2 User Manual.

If a display similar to Figure 14 cannot be produced, SEEK SPECIALIST

ADVICE before proceeding.

8. Save the data file.

9. Check all probes to be used in the session (Repeat steps 4-8)

Inspection Procedure
1. Mark up the inspection area. For flat plate and T-butt geometries, mark positions at 100mm intervals
from left hand plate edge using washable marker pen or chalk. For cylindrical geometries, mark
clock positions or 100mm markers as appropriate.

1378.DOC Rev 12 Page 28

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

2. PC operator to inform probe operator of scan required and which probe to use. Connect the probe
and confirm the probe serial number.
3. Select the required probe configuration.
4. Create a new file for the inspection data.
• Choose a filename in the form "XXXXXXXa"
where: XXXXXXX = Identifying code for location to be inspected (up to 7 characters).
a = Inspection number/letter for that location (e.g. a for first inspection, b for second inspection
etc.).
When choosing a data filename, make sure the file is uniquely identifiable and contains a reference
to the component under test.
Note: The individual scans within the inspection are described by pages within the file, in numerical
order. A separate recording sheet or log-book must record the scan made on each page (i.e. the weld
toe/bead inspected and the direction of the probe 'C' or 'A').
5. Set the range of the numbered marker lines (or clock-points) to be covered in the scan (start, end and
direction)
While recording data, reference markers can be entered which are called CLOCK POINTS. These
are lines labelled with a number between 1 and 99. The numbers wrap around at the end point so
that they can be used to mark positions when scanning around a tubular geometry. The default end
point is 12, appropriate for conventional clock positions around a circle, however the markers can
just as easily be used to scale mark a linear geometry. e.g. each marker may represent 100mm
increments.
6. Probe operator to place probe at beginning of scan, and confirm probe direction (Probe marker A or
C leading) to PC operator. When A is leading, the loop produced by a defect in the butterfly plot
should be traced in an anti-clockwise direction, when C is leading it should be clockwise.
7. Start data collection and begin scanning. Note that because data collection can be started and
stopped from the PC, the Amigo or (in some cases) the probe, the detailed procedure for co-
ordinating data collection in 2-man operation must be agreed beforehand. Scan speed approx.
50mm/sec, but if this needs to be reduced for operational reasons, the data sampling rate should be
reduced accordingly.
8. Probe operator to report when centre of probe passes marker. Probe operator or PC operator to log
this position. (In 2-man operation, it may be more convenient for the probe operator to log these
points using buttons on the Amigo or the probe). The Probe operator should report visual indications
such as seam welds, weld run overlap, weld spatter or grinding marks when these are encountered.
The PC operator may add an un-numbered marker to log these.
9. Stop scanning and halt data collection.
10. Inform probe operator that probe may be removed from component. Change scan speed (data
collection rate) if necessary.
11. Study the data, looking for any defect indications.
For defect detection refer to “INTERPRETATION OF ACFM SIGNALS” on page 20.
Any defect signal is to be noted in the report sheets by the operator and the area marked by the probe
operator.
12. Continue scanning with a minimum of one clock position, or 100mm overlap on each scan (e.g.
clock positions 11-3, 2-6, 5-9, 8-12).
Repeat steps 5 to 11 until scan area is covered.
When defects are detected, additional scans are necessary in order to facilitate sizing. These are:
i) a slow scan to cover all defects within a region.

1378.DOC Rev 12 Page 29

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

ii) a scan or series of scans to locate the ACFM crack ends. In these scans the PC operator will
ask the probe operator to move the probe so that it coincides with the peaks and troughs of
the Bz signals (usually easiest to observe on the butterfly plot). The marking will normally
be aligned directly with the probe centre. Alternatively, if the access does not allow this, the
trailing edge of the probe can be marked first with the position of the centre of the probe
marked later.
13. Save the collected data in the file selected at the beginning of this session.
14. Probe operator should place a ruler or tape on the specimen and read off crack end positions to the
PC operator. The PC operator must record this “Bz crack length” and the distance to a datum on the
data sheet.
15. Repeat steps 1 to 14 for all inspection areas on the specimen.
16. Crack depth sizing can be carried out immediately, or at a later date, in accordance with the
procedure described in "Crack Sizing" on page 23.
17. Close QFM, shut down Windows and turn off the Amigo.

End Of Session Ops-Check And Backup Procedure

Post-Session Ops-Check
To ensure that the system was functioning correctly during the session, the ops check should be repeated on
the appropriate sample at the end of the shift using the procedure outlined in “Operations Check” on page 27,
steps 3 to 9.

Backup Procedure
1. Mark back-up medium (e.g. floppy disc, magneto-optical disc, CD-R) with files to be copied.
2. Put back-up medium in appropriate drive
3. Back-Up all files created in the inspection shift.

Exit QFM3 and check backup

1. Exit QFM:
2. Check files are stored on the backup medium using (e.g.) Windows Explorer
3. Remove the backup medium from drive and write-protect (if applicable).
Switch off all power.

1378.DOC Rev 12 Page 30

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

APPENDIX A

Probe Specifications

Standard Weld Probe

The Weld Probe is used whenever possible for the routine ACFM inspection of welds, plate and tubulars.
Other probes are available for inspection in areas of restricted access or areas close to plate edges.
The probe contains a sensor coil pair consisting of a Bx and Bz coil wound concentrically. The bottom face is
fitted with two feet which, together with the wear-resistant probe nose housing the sensors, give a stable
tripod arrangement for ease of scanning.
Probes can be supplied with probe extension cables or longer fitted cables. Details on request.

The standard ACFM weld probe

1378.DOC Rev 12 Page 31

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Mini-Pencil Probe
Sub-Types Straight Entry Cable
Right Angle Entry Cable, parallel nose
Right Angle Entry Cable, transverse nose
The Mini Pencil Probe maintains the sensitivity of the normal Weld Probes. They have the advantage of
being able to access restricted areas and areas of weld associated with rat holes in structures. They are less
prone to signal variations near plate edges, but are less accurate for depth sizing deep defects (>5mm). The
parallel nose type is ideal for inspecting through rat-holes, or for longitudinal cracks inside pipes. The
transverse nose is ideal for circumferential cracks inside pipes.
Probes can be supplied with probe extension cables or longer fitted cables. Details on request.

Straight entry miniature pencil probe

1378.DOC Rev 12 Page 32

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Right Angle cable, parallel nose miniature pencil probe

Right Angle cable, transverse nose miniature pencil probe

1378.DOC Rev 12 Page 33

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Micro-Pencil Probe
Sub-Types Straight Entry Cable
Right Angle Entry Cable, parallel nose
Right Angle Entry Cable, transverse nose
The Micro Pencil Probes are similar to the Mini Pencil Probes but are manufactured with high sensitivity
coils for the detection and measurement of shallow defects, less than 1mm deep.
The construction of the probe necessitates the search coils not being coincident, which makes the butterfly
display distorted, because the Bx and Bz signals are not in line as seen with normal ACFM probes.

Straight entry micro pencil probe

1378.DOC Rev 12 Page 34

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

Right Angle cable, parallel nose micro pencil probe

Right Angle cable, transverse nose micro pencil probe

1378.DOC Rev 12 Page 35

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

AMIGO Underwater Weld Probe

The AMIGO Underwater Weld Probe is designed for subsea inspection of welds, plate and tubulars. The
probe is sealed and fitted with a 50m cable. The probe electronics are encapsulated in a bulge in the cable
about 30cm from the probe body.
If the probe is to be used in an environment where sharp marine growth could easily damage the probe cable,
it is recommended that the cable be protected using additional sheathing.

54.0

58.0

50.0

The underwater ACFM weld probe

AMIGO Underwater Mini-Pencil Probe

Sub-Types Straight Entry Cable
Right Angle Entry Cable, parallel nose
Right Angle Entry Cable, transverse nose
The Mini Pencil Probe maintains the sensitivity of the normal Weld Probes. They have the advantage of
being able to access restricted areas and areas of weld associated with rat holes in structures. They are less
prone to signal variations near plate edges, but are less accurate for depth sizing deep defects (>5mm). The
parallel nose type is ideal for inspecting through rat-holes, or for longitudinal cracks inside pipes. The
transverse nose is ideal for circumferential cracks inside pipes.

1378.DOC Rev 12 Page 36

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

STRAIGHT CABLE ENTRY RIGHT ANGLE CABLE ENTRY (PARALLEL NOSE)

RIGHT ANGLE CABLE ENTRY (TRANSVERSE NOSE)

AMIGO Underwater Micro-Pencil Probe

Sub-Types Straight Entry Cable
Right Angle Entry Cable, parallel nose
Right Angle Entry Cable, transverse nose
The Micro Pencil Probes are similar to the Mini Pencil Probes but are manufactured with high sensitivity
coils for the detection and measurement of shallow defects, less than 1mm deep.
The construction of the probe necessitates the search coils not being coincident, which makes the butterfly
display distorted, because the Bx and Bz signals are not in line as seen with normal ACFM probes.

1378.DOC Rev 12 Page 37

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

STRAIGHT CABLE ENTRY RIGHT ANGLE CABLE ENTRY (PARALLEL NOSE)

RIGHT ANGLE CABLE ENTRY (TRANSVERSE NOSE)

Underwater Pencil Probe

The Underwater Pencil Probes are designed to fit into grind repairs. They are built to be factory serviceable
but should not be disassembled on site.

1378.DOC Rev 12 Page 38

ACFM Inspection Procedure QFM Vn. 3.x and Amigo

The underwater pencil probe

Other Probe Types

Other probe types may be introduced as part of TSC's continuing policy of product development. If a suitable
probe for a particular application is not shown here, contact TSC.

Purpose of Probe Files

The aims of a probe file are:
a) To maintain a consistent detection sensitivity between inspections, and between probes.
b) To maintain sizing accuracy
The first aim is satisfied by the probe file containing graph scalings to give a standard size butterfly loop from
a given defect.
The second aim is satisfied by the probe file containing calculated values for inherent coil lift-off and
sensitivity.
If, after checking instrument settings etc., it is found that the issued probe file is unsuitable for a particular
application (for example the inspection is not on ferritic steel or the readings are saturating), refer to TSC. If
approval is given to create a new probe file on site, TSC will issue the appropriate procedure. It should be
noted that to produce a new probe file, an ops check plate in the same material as that under inspection will
be needed, containing a slot (preferably 50mm long by 5mm deep).

1378.DOC Rev 12 Page 39

 APPENDIX F
Ultrasonic Phased Array Results

15558/6/06
Copyright © 2006, TWI Ltd
Scan 1 from 316-01

Scan 2 from 316-01

F1

15558/6/06
Copyright © 2006, TWI Ltd
Scan 3 from 316-01

Scan 4 from 316-01

F2

15558/6/06
Copyright © 2006, TWI Ltd
Scan 1 from 316-02

Scan 2 from 316-02

F3

15558/6/06
Copyright © 2006, TWI Ltd
Scan 3 from 316-
02

Scan 4 from 316-02

F4

15558/6/06
Copyright © 2006, TWI Ltd
Scan 1 from 2205-2

Scan 2 from 2205-2

F5

15558/6/06
Copyright © 2006, TWI Ltd
Scan 3 from 2205-2

Scan 4 from 2205-2

F6

15558/6/06
Copyright © 2006, TWI Ltd
Scan 5 from2205-
2

Scan 6 from 2205-2

F7

15558/6/06
Copyright © 2006, TWI Ltd
Scan 1 from 2205-3

Scan 2 from 2205-3

F8

15558/6/06
Copyright © 2006, TWI Ltd
Scan 1 from 2205-4

Scan 2 from 2205-4

F9

15558/6/06
Copyright © 2006, TWI Ltd
Scan 3 from 2205-4

Scan 4 from 2205-4

F10

15558/6/06
Copyright © 2006, TWI Ltd
Scan 5 from 2205-4

Scan 1 from Duplex pipe 01

F11

15558/6/06
Copyright © 2006, TWI Ltd
Scan 2 from duplex pipe 01

Scan 1 from duplex pipe 02

F12

15558/6/06
Copyright © 2006, TWI Ltd
Scan 2 from duplex pipe 02

Scan 3 from duplex pipe 02

F13

15558/6/06
Copyright © 2006, TWI Ltd
UT Ins. CRB 1, Issue 1, Rev.1 June 2006

TWI NDT TECHNOLOGY GROUP

Instruction Reference: PAUT/KSH 1

Component Identification: Cracked Test Blocks 2205/2-5 and 316ss 1 &

2, Pipes 1 &2

Written by: K S Hayward

Qualifications: CSWIP Level 2
Date: January 2006

Approved by: C R Bird

Qualifications: PCN Level 3
Date: January 2006

Company Name: TWI Ltd

Action to be taken
On location of defects: Store data and report

Page 1 of 4
UT Ins. CRB 1, Issue 1, Rev.1 June 2006

TECHNIQUE SHEET FOR THE INSPECTION OF 12mm THICK

DUPLEX and 316 STAINLESS STEEL PLATE AND PIPE

PURPOSE OF TEST

To provide a study of the detection of stress corrosion cracks in plate samples of duplex and
stainless steel welds.

The position of the cracks on the surface is known.

INSPECTION PROCEDURE

Equipment

Phased Array Probe

7MHz, 32 element, element pitch 0.6mm, 10mm wide 33° wedge, rexolite wedge material.

Calibration Blocks

Calibration blocks in accordance with BS EN 12223 and BS EN 27963 plus calibration block
CRB1 Fig.1 with a line of vertical 3mm side drilled holes (SDH) and a 45° planar reflector.

20.00

100mm

Ø3.00

Fig.1 Calibration block

Instrument

Tomoscan Focus phased array ultrasonic instrument with the following specification:

32 active channels in an array of 128 channels.

A-scan display capable of time-base and amplifier linearity performance in accordance with
the requirements of BS EN 12668.

Validated and approved software for data collection analysis using the specified instrument.

Page 2 of 4
UT Ins. CRB 1, Issue 1, Rev.1 June 2006

Couplant

Ultrasonic couplant Ultragel II

Manipulator

Single axis manipulator Phoenix ISL string encoder.

Personnel

All personnel working in accordance with this procedure shall be qualified as a minimum to
EN 473 ultrasonic practitioner level 2. Additionally all personnel to hold a current level 2
CSWIP in manual phased array ultrasonic applications with R/D-Tech phased array
equipment (Omniscan, Focus or Tomoscan III).

Procedure Pre-requisites

Check time-base and amplifier linearity in accordance with BS EN 12668. Perform phased
array system element check to demonstrate that all elements are functioning. If more than
three elements are not functioning the probe shall be rejected.

Check material surface finish.

Pre-clean component surface to remove all traces of poorly adhered paint.

Perform 0° lamination check.

Safety Precautions

All operators will wear the appropriate safety clothing and protective footwear.
All electrical equipment will carry a valid electrical certificate.
All chemicals to be used shall have relevant Heath and Safety Data sheets available.

Calibration and Scanning Sensitivity

Range: 40mm in steel

Reference signal: 3mm diameter side drilled hole
Reference signal amplitude: 80% full screen height (FSH)
Scan gain for defect detection: Reference gain +10dB

Equipment Set-up Parameters

Beam angles 40° to 70° at 1.0° intervals

Focal depth 24mm (full skip)

Ultrasonic filters: Low pass 2MHz

High pass 10MHz
Smoothing 4MHz

Page 3 of 4
UT Ins. CRB 1, Issue 1, Rev.1 June 2006

Digitisation frequency: 12.5MHz

Data Collection procedure

1. Calibrate the time-base for 160mm

2. Check transfer loss using a V path measurement on an equivalent thickness test block
3. Scan the probe across the plate to cover the cracked area as appropriate
4. Scan probe using the single axis string encoder with a data recording increment of
2mm.
5. Ensure that the probe scan speed is no greater than the data recording threshold of the
phased array instrument.

Data Analysis Requirements

Visual indications of observed cracks as required, identified with the surface indication if
possible.

Reporting

All crack-like flaws to be reported.

A report of the inspection will be produced containing both plan and end views of the weld
and showing for each scan.

Post Test Procedure

All traces of ultrasonic couplant will be removed from the test surface and the finish restored
to its original state.

Non-Compliance

Should the operator be unable to follow this instruction for any reason then guidance will be
sought from his supervisor before proceeding.

Page 4 of 4
 APPENDIX G
Pulsed Eddy Current Test Report

15558/6/06
Copyright © 2006, TWI Ltd
 Report
Pulsed Eddy Current
By University of Huddersfield

Contents

1 Introduction ___________________________________________________________________________________ 2

2 Experimental Procedures_________________________________________________________________________ 3

3 Results 1 (Top Side) _____________________________________________________________________________ 4

3.1 Sample 1 (with fewer visible defects) ____________________________________________________________ 4

3.1.1 1st Scanning Line (1a) ______________________________________________________________________ 4
3.1.2 2nd Scanning Line (1b) _____________________________________________________________________ 4
3.1.3 3rd Scanning Line (1c) _____________________________________________________________________ 5
3.2 Sample 2 (with more visible defects) _____________________________________________________________ 5
3.2.1 1st Scanning line (2a) _______________________________________________________________________ 5
3.2.2 2nd Scanning Line (2b)______________________________________________________________________ 6
3.2.3 3rd Scanning Line (2c) ______________________________________________________________________ 6
3.3 Combined Results ___________________________________________________________________________ 6

4 Results 2 (Bottom Side Scanning) __________________________________________________________________ 8

4.1 Sample 1 (with fewer visible defects on the other side)_______________________________________________ 8

4.2 Sample 2 (with more visible defects on the other side) _______________________________________________ 9

5 Discussion ____________________________________________________________________________________ 10

G1

15558/6/06
Copyright © 2006, TWI Ltd
1 Introduction

We have been provided with 2 samples of similar types:

1. Coded 2205-3. This has fewer visible cracks in the weld area. The cracks are oriented
roughly parallel to the weld line. In this report, this sample is referred as Sample 1.
2. Coded 2205-5. This has more visible cracks. Will be referred as Sample 2.

Each sample has dimensions of 34.0 cm x 15.5 cm x 1.25 cm. Through this preliminary
investigation, the potential performance of PEC is studied.

35

40

35

Figure 1 Sample 1 with fewer cracks

40
a

40
b

40

Figure 2 Sample 2 with more visible cracks

G2

15558/6/06
Copyright © 2006, TWI Ltd
2 Experimental Procedures
For each sample, scanning is carried out parallel to the length of each sample, and
perpendicular to the weld line. Three scans are carried out for each sample, i.e. scans a, b, and
c as shown in figures 1 and 2. Please not different positions of the lines between sample 1 and
sample 2. The scanning will be done on both sides of each sample, the side in black is referred
as the bottom side, and the other side is taken as the top side.
The approximate centre of the weld line is taken as x = 0. The scans are carried out manually,
data is collected every 5 mm, from x = -50 to x = +50 mm, where negative distances are on the
left hand side to x = 0. Between x=-10 and x=+10, data is collected every 1 mm.
For preliminary observation, peak values and peak arrival times against distance will be
presented. Signals with peak values less than 0.6 after normalization are considered too small
for consideration and hence their peak arrival time will be ignored and made equal to 0.
The probe to be used has a multilayered cylindrical excitation coil of 40 turns with ID of 7 mm,
OD of 9 mm, and height of 0.5 mm. The sensing device is a GMR. The amplitude of the
excitation current used is 1.5A.

G3

15558/6/06
Copyright © 2006, TWI Ltd
3 Results 1 (Top Side)
3.1 Sample 1 (with fewer visible defects)

3.1.1 1st Scanning Line (1a)

-3
x 10
4

) 3
br
a(
e 2
ul
a
V
k 1
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

The signatures of both the peak arrival time and height suggest that there may be at least 2
cracks existing along the scanned line. Visual observation suggests only 1 crack is present.

3.1.2 2nd Scanning Line (1b)

0.015

)
br 0.01
a(
e
ul
a 0.005
V
k
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

These results suggest that there are some surface defects that are very small in size and are
located mainly on the left to the weld line. The defects are not obvious for visual observation.
The large peak values take place between x=-50 and -40 mm can be due to tilting effects or
some ‘variations’ in the material. However, these could be easily filtered out as their peak arrival
times are not similar to those of surface cracks.
G4

15558/6/06
Copyright © 2006, TWI Ltd
3.1.3 3rd Scanning Line (1c)
-3
x 10
4

) 3
br
a(
e 2
ul
a
V
k 1
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

3.2 Sample 2 (with more visible defects)

3.2.1 1st Scanning line (2a)

-3
x 10
8

) 6
br
a(
e 4
ul
a
V
k 2
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

Clearly 2 separate lines of cracks are suggested by the resulting graphs.

G5

15558/6/06
Copyright © 2006, TWI Ltd
3.2.2 2nd Scanning Line (2b)

0.01

) 0.008
br
a( 0.006
e
ul
a 0.004
V
k 0.002
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.25

) 0.2
s
m
( 0.15
e
m
i 0.1
T
k
a 0.05
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

The resulting graphs suggest that 3 cracks may take place that are crossed by the scanned line.

3.2.3 3rd Scanning Line (2c)

-3
x 10
8

) 6
br
a(
e 4
ul
a
V
k 2
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

Clearly 2 separate lines of cracks are suggested by the resulting graphs.

3.3 Combined Results

Using time windowing we could filter out the peak values that are not coming from surface
cracks. Below is the zoomed-in graphs for better viewing of the areas closer to the x = 0 mm.
The graphs also illustrate comparative depths of the detected cracks.

G6

15558/6/06
Copyright © 2006, TWI Ltd
 0.01 0.25

0.009
1a
0.008 1b 0.2
1c
0.007
2a
) )
b 2b
r 0.006 s 0.15
a( 2c m
(
e 0.005 e
ul m
a i
V 0.004 T 0.1
k k 1a
a a
e 0.003 e 1b
P P 1c
0.002 0.05 2a
2b
0.001
2c

0 0
-15 -10 -5 0 5 10 15 -15 -10 -5 0 5 10 15
Distance (mm) Distance (mm)

G7

15558/6/06
Copyright © 2006, TWI Ltd
4 Results 2 (Bottom Side Scanning)
For the bottom side only one scanning done for each sample.

4.1 Sample 1 (with fewer visible defects on the other side)

-4
x 10 Normalised Differential Signal
8

) 6
br
a(
e 4
ul
a
V
k 2
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

G8

15558/6/06
Copyright © 2006, TWI Ltd
4.2 Sample 2 (with more visible defects on the other side)

-3
x 10 Normalised Differential Signal
1

) 0.8
br
a( 0.6
e
ul 0.4
a
V
k 0.2
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50

0.2

) 0.15
s
m
(
e 0.1
m
i
T
k 0.05
a
e
P
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
X (mm)

G9

15558/6/06
Copyright © 2006, TWI Ltd
5 Discussion
- Top side scanning
o The results show that the technique could pick up the cracks.
o Cracks closed to each other distanced by as small as 2 mm seem to be separated
successfully using the probe
o Although no defects are visible on the scanning line 1b, but the results suggest
that some defect is present.
o The results also present the relative magnitude of the cracks. The cracks on
Sample 2 seem to be more severe than cracks on the other sample
- Bottom side scanning
o The recorded peak values are relatively low. It is apparent that they are
inadequate to indicate any possible discontinuities.

G10

15558/6/06
Copyright © 2006, TWI Ltd
 APPENDIX H
EMAT and Laser Ultrasound Test Report

15558/6/06
Copyright © 2006, TWI Ltd
 Report

The Ultrasound Group

Department of Physics
University of Warwick

EMAT and laser ultrasound

measurements of some samples
provided by TWI
Dec 2005 / Jan 2006
Rachel S Edwards, Y. (Charley) Fan and Steve Dixon
University of Warwick

In confidence; not to be distributed

No part of this report may be copied or otherwise used without

the express permission of the authors, including all text, figures
and results.

© University of Warwick, March 2006

1. Introduction
Photographs of the samples along with descriptions of the scan positions are given in
the PEC probe report.
1. 2205-3; this has fewer visible cracks in the weld area. The cracks are
oriented roughly parallel to the weld line (labelled sample 1).
2. 2205-5; this has more visible cracks (labelled sample 2).
The potential of crack detection on this type of sample using Rayleigh waves was
investigated.

Rayleigh waves were generated and detected using a pair of EMATs with a fixed
separation and frequency content around 200 kHz.

Firstly, a scan over the weld area was done. The EMATs in their holder (fixed separation
of 150 mm) were placed at the top of the sample with the generation and detection
EMATs on either side of the defects, and scanned together over the weld area. This was
done for both samples, for the top and bottom sides, with a comparison scan done with
both EMATs over a defect free region.

Secondly, scans were done along lines a, b and c, as used for PEC testing, for the top
and bottom sides of each sample. In this case the EMATs were held at a fixed
separation of 124 mm. The approximate centre of the weld line is at x = 0, with the scans
done over the range permitted by the limited sample size.

Finally, some scans were performed using laser generation and detection, for higher
spatial resolution and higher frequency content of the signal.

2. Results (EMAT-EMAT measurements)

Top side scanning

We show in figure 1 typical results from scanning the top of sample 2, which contains
several cracks. The B-scan has time on the x-axis and scan position on y. With the
EMATs close to the sample edge (around x = −30 mm) we can see reflections from the
sample edge; with the EMATs close to the crack (around x = −3 mm) we can also see
reflections from the defects.
 Figure 1: B-scan on sample 2, scan c

We use changes in the peak to peak signal amplitude and frequency content of the
Rayleigh wave to gauge the depth of defects in thick samples. In figure 1, the feature
around 48 µs is affected by the defects, and changes in its amplitude are shown in figure
2.
Normalised amplitude (arb. units)

1.6 (b)
1.4

1.2

1.0

0.8

0.6
Top
0.4 2a
2b
0.2
2c
0.0
-30 -20 -10 0 10 20
Scan distance (mm)
Figure 2: Normalised peak to peak amplitudes for scans on top of sample 2

At a crack we may see an enhancement, as the direct Rayleigh wave interferes with that
reflected by the crack, followed by a drop in signal as the crack blocks some of the signal
from reaching the receive EMAT. Both of these features can be seen in figure 2; an
enhancement for each scan around −8 to −2 mm, and a drop in signal for positive x
values. The enhancement gives the crack position (in fact, several can be seen on the
scans, for example the feature around x=+5 mm on scans 2a and 2c), and the depth can
be gauged from the transmitted signal amplitude.

Similar results are found on sample 1; however, the cracking is less severe and may be
better suited to the use of higher frequencies.
Bottom side scanning

It is interesting to see a useable variation in signal amplitude when scanning from the
opposite side of the sample to the cracks, and it would be good to test this on a
calibration sample.
4
Top
Base

Relative depth (mm)

3

0
20 40 60 80 100 120
Scan distance (mm)
Figure 3: Relative depths measured on scan of weld from bottom of sample 1

On scanning the weld area of sample 1, we see a variation in transmitted signal

amplitude which can be associated with a change in depth of the cracks between the
two EMATs, and the variation is similar for scans from the top and bottom of the sample.
For reasons discussed later, the depths given in figure 3 are intended as an indication of
the depth variation along the weld, and not as the actual depths. However, from both
sides we are measuring two main regions of cracking, as is apparent when studying the
top of sample 1 by eye.

On sample 2, similar results to those shown in figure 1 are found when scanning from
the sample base.

3. Results (laser ultrasound experiments)

We have done scans using laser generation and detection of surface ultrasonic waves,
which gives higher frequency content and higher special resolution. Testing was done
for scans 1a, b and c, and 2b and c, from the top only. Figure 4 shows the peak to peak
amplitude of the Rayleigh wave signal. We can clearly see an enhancement followed by
a drop in signal at x = −6.5, followed by a plateau and another enhancement and drop at
x = −0.5 as the generator passes over the second crack.
 1.2

Normalised amplitude (arb. units)

1.0

0.8

0.6

0.4

0.2

0.0
-15 -10 -5 0 5 10
Scale (mm)
Figure 4: Normalised signal amplitudes from laser scan of 2b

4. Discussion
• Sample size – the samples were too small to accommodate the EMAT trolley, so
manual scans were necessary. Also, the normal EMAT-EMAT separation had to be
shortened to 124 mm to allow scanning over the crack.
o A larger sample size would allow more reliable measurements
• The sample thickness is less than the main Rayleigh wavelength when using these
EMATs, which causes difficulties in accurately windowing the correct surface wave to
use for depth gauging, in particular when FFT-ing the signal for studying the
frequency content. As we are not working strictly in the Rayleigh regime the depth
calibration done on thick samples will not be accurate.
o A calibration sample with known defects would be useful to study the effect of
working with thicknesses less than the Rayleigh wavelength.
o Alternatively, higher frequencies could be used, as in the laser experiments
which show very promising results. However, this may lead to problems in
detecting cracks from the bottom of the sample.

The EMAT-EMAT technique can detect the surface defects, and, given the correct
calibration and windowing for finding the frequency content, should be able to gauge the
depth of the deepest defect present between the two EMATs. Higher spatial resolution
would be beneficial, which can be gained from using smaller EMATs (currently being
developed) or laser techniques.

Proposed future work

Investigations of a larger sample, to allow easier scanning, would be beneficial.
Calibration measurements on a sample containing well known defects would enable the
benefits of using this technique to detect this sort of cracking to be thoroughly
investigated, and a full solution proposed. The EMAT design could be modified to allow
higher special resolution and higher frequencies to be used when scanning the samples,
and potential designs of EMAT and holder would be investigated.
 APPENDIX I
Shearography Test Report

15558/6/06
Copyright © 2006, TWI Ltd
September 2005
Reiko Takahashi

Shearography Measurement for Stress Corrosion Cracking Plates

1. Introduction

Within various non-destructive testing techniques, laser shearography was also

selected as one of surface inspection methods to detect stress corrosion cracking on
austenitic and duplex stainless steel plates.

2. Principle of shearography measurement

Shearography is a large area optical inspection system that utilises laser light to detect
very slight surface deformations due to subsurface discontinuities. It is a full field non
contact method. It uses the interference of laser light to detect surface displacements.
These surface deformations are generated by subjecting the part to a controlled
stressing mechanism such as a vacuum, vibration heat or mechanical force.

Shearography employs a single beam of expanded laser light, reflected off the
specimen. The CCD camera then produces a pair of laterally sheared images in the
image point on the image into two points in the image as shown in Fig 1. The two
overlapped portions of the sheared images interfere and produce a speckle pattern.
When the object is deformed, the speckle pattern is slightly modified. Comparing the
two speckle pattern approximately represents the derivative of displacement with
respect to the shearing direction. Shearography measures the gradient of the
deformation, not the deformation itself as holography does. This difference results in
shearography being much less sensitive to external environmental factors and can be
applicable to production and field environments.

Object

Laser

Image CCD Camera

Fig.1 Shearography set-up

I1

15558/6/06
Copyright © 2006, TWI Ltd
NDT of shearography is mainly used for composite materials, like honeycomb
structures (as seen Fig.2) but the application for metals are also under investigation.
Software is well developed and the interpretation is simple– to find a butterfly pattern.

Butterfly pattern

Shearographic
result

Fig. 2 Example of shearography output

3. Test set-up

Specimen
Jig

Laser

CCD Camera

Fig. 3 Shearography measurement set-up

All the equipment was set on the optical bench to avoid any relative movement.
Various stressing methods were discussed prior to the measurement. For a metal
specimen, the heat loading is not good since the heat will transmit so quickly and
never stay at the area of interest, which will not help to generate surface deformation.
The vacuum excitation method was tried but the metal specimen was too stiff to
generate local strain around the cracks. Then the 3 point bending jig was selected to
apply mechanical stress from the back of the specimen. As seen in Fig. 3, the laser
I2

15558/6/06
Copyright © 2006, TWI Ltd
light illuminates the surface of the specimen. The shear vector was set perpendicular
to the bending direction.

4. Result

Three samples were tested.

Sample A: A plate with large cracks
Sample B: A plate with fine cracks
Sample C: A plate with no crack

Crack indication was observed only on the sample A, which can be seen in Fig.4.
Three cracks were found.

Fig. 4 Shearography result

5. Conclusion

• Shearography can be used to detect the stress corrosion cracks.

• Shearography is able to detect large cracks, approximately more than 3mm deep.
• The stressing method must be properly selected for the best result.

I3

15558/6/06
Copyright © 2006, TWI Ltd
6. Discussion

The result proved that it was possible to detect stress corrosion cracks on austenitic
and duplex stainless steel plates by shearography. It is also possible to measure
approximate size of the cracks if you set parameters properly in the software. This
will help to determine engineering criteria for the selected application. For
shearography measurement, the selection of stressing method is vital. Also it is
important to know the target defect size to determine the ideal level of the load.

I4

15558/6/06
Copyright © 2006, TWI Ltd