Beruflich Dokumente
Kultur Dokumente
CITATIONS READS
0 20
3 authors, including:
Dheeraj P R
Indian Institute of Technology Kharagpur
3 PUBLICATIONS 0 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Dheeraj P R on 30 September 2017.
Ultrasonic imaging examination methods such as phased array ultrasonic testing (PAUT) and time-of-flight diffraction
(TOFD) are effective methods for the volumetric examination of welds. This is true where access from both sides is possible
for scanning; however, for welds with single-sided access, such as welds at the elbow, flange and cross-section of pipes,
limited scanning compromises the detection of lack of fusion (LOF) on the bevel opposite to the probe side. As a common
industrial practice, single-sided access ultrasonic weld examination for welds with a crown (as-welded condition) using
a linear phased array probe employs a second leg ultrasonic beam examination that utilises the tip diffraction signals
for LOF detection or a third leg ultrasonic beam examination that utilises the specular reflection signal for LOF detection.
This paper considers a linear array probe as conventional phased array technology and a dual matrix array (DMA) probe
as advanced phased array technology. The objective of this paper is to evaluate the probability of LOF detection through
conventional and advanced phased array technology in single-sided access carbon steel weld examination through a
case study using simulation software and practical examination on welded specimens. The probability of LOF detection is
considered based on the amplitude, signal-to-noise ratio (SNR) and sizing of the ultrasonic signal response from the LOF.
Keywords: ultrasonic, conventional phased array technology, linear array probe, advanced phased array technology, DMA
probe, single-sided access welds, simulation, focal law, probability of detection.
Simulation Practical
3. Specimen selection
Figure 1. Typical probe placement with a second leg ultrasonic
Welded specimens of two different thicknesses, 12.5 mm and
beam to cover the far-side bevel
20 mm, with weld crowns are chosen for this study. As LOF is the
most commonly missed defect in single-sided access ultrasonic
As the bevel face is close to parallel for the second leg of the weld examination, it has been chosen as the target reflector. Each
beam, there will not be any possible reflected echo from the LOF. specimen contains a natural LOF with a specific length, height and
However, there is a possibility of weak tip-diffracted echo signals position in the weld. All the LOFs are chosen to be on one side of the
from the extreme tips of the LOF, which requires a considerable bevel and this bevel side is termed as the skew 90 side. The opposite
increase in dB (gain) level for tip detection. bevel side, where there is no LOF, is termed as the skew 270 side. A
The third leg of the ultrasonic beam from a linear phased array typical skew 90 side and 270 side is illustrated in Figure 5.
probe is used to cover the far-side bevel region. A typical probe
placement with a third leg to cover the far-side bevel is illustrated
in Figure 2.
For conventional phased array technology, Probe 1, a linear
array 5 MHz probe of 1.0 mm pitch, is used. For advanced phased
array technology, Probe 2, a DMA 4 MHz probe of 1.0 mm pitch,
is used. Figure 3 illustrates the wedges (natural angle 55º, velocity
2330 m/s) used in this study. Equipment with 32 pulsers and 128 Figure 5. Typical skew
channels is used.
The dimensions of the LOFs are evaluated using conventional
phased array technology (scan plans 1 and 5), which utilises linear
phased array probe placement at skew 90 (the same side as the LOF,
as of two-sided weld examination) at reference dB and listed in Table
1. Reference dB is considered as the dB level at which time-corrected
gain (TCG) is drawn based on an 80% full screen height (FSH) echo
response from a 2.5 mm SDH of 38 mm length at different depths.
4. Scan plan
Figure 2. Typical probe placement with a third leg ultrasonic beam
to cover the far-side bevel The conclusion from a previous study[1] has been used to generate
the most appropriate scan plans for phased array examination.
Cross-section view
The scan plans for a 12.5 mm-thick weld specimen are provided a linear phased array probe. Scan plan 8 provides the first leg of
in Table 2. Scan plan 1 provides the second leg of the ultrasonic the ultrasonic beam for examination from skew 270 using a DMA
beam for examination from skew 90 using a linear phased array probe.
probe. Scan plan 2 provides the second leg of the ultrasonic beam
for examination from skew 270 with the tip diffraction technique
using a linear phased array probe. Scan plan 3 provides the third
5. Objectives of the case study
leg of the ultrasonic beam for examination from skew 270 using The objectives of the case study are:
a linear phased array probe. Scan plan 4 provides the first leg of l To perform a simulation and practical study on the
the ultrasonic beam for examination from skew 270 using a DMA welded specimens as described in Section 3 of this paper to
probe. detect skew 90 side LOFs through the use of single-sided access
The scan plans for a 20 mm-thick weld specimen are provided ultrasonic weld examinations, as described in Section 2 of
in Table 3. Scan plan 5 provides the second leg of the ultrasonic this paper, with the scan plan as provided in Section 4 of this
beam for examination from skew 90 using a linear phased array document.
probe. Scan plan 6 provides the second leg of the ultrasonic beam l To tabulate the results of each selected single-sided access
for examination from skew 270 with the tip diffraction technique ultrasonic weld examination. The result includes amplitude in
using a linear phased array probe. Scan plan 7 provides the third %FSH, SNR and height of the LOFs. The tabulated results are the
leg of the ultrasonic beam for examination from skew 270 using practical results obtained from PA equipment.
6. Case study 1: 12.5 mm with probe placement at skew 270 to detect the tip-diffracted signals
of the LOF. The S-scan images and observed results (amplitude,
6.1 Conventional phased array technology SNR and height) of scan plan 2 over LOF1 and LOF2 are tabulated in
(scan plan 1) Tables 5 and 6, respectively.
At the reference dB, it is observed that the LOFs are almost
Scan plan 1 utilises the second leg of the ultrasonic beam from a missed, as the tip-diffracted signals are much weaker. The reference
linear phased array probe with probe placement at skew 90. The dB is increased gradually until the tips of the LOF are accurately
S-scan images and observed results (amplitude, SNR and height) detected. LOF1 is accurately detected at +25 dB and LOF2 is
of scan plan 1 over the two natural LOFs are tabulated in Table 4. accurately detected at +26 dB. However, the additional dB depends
Sizing is performed using the 6 dB amplitude drop method. These on the nature of the LOF tip and may vary considerably from case
observed results are considered as a reference for comparing the to case.
results of the other scan plans: 2, 3 and 4. Sizing is performed by measuring the distance between the tip-
diffracted signals. It is observed that an increase in dB for detection
6.2 Conventional phased array technology comes at the cost of an increased noise level, which is undesirable
(scan plan 2) above a certain extent. A greatly reduced SNR and noise signals
mixed with defect signals will create confusion for the interpreter.
Scan plan 2 utilises the tip diffraction technique, which uses the This tip diffraction technique can be applied on a case-by-case basis
second leg of the ultrasonic beam from a linear phased array probe with great caution in respect of a desirable SNR.
Table 4. S-scan image and results of scan plan 1 for LOF1 (left) and LOF2 (right)
LOF1 LOF2
Practical Simulation Practical Simulation
LOF1
Reference dB +25 dB
Practical Simulation Practical Simulation
LOF2
Reference dB +26 dB
Practical Simulation Practical Simulation
6.3 Conventional phased array technology additional +12 dB is required above the reference dB.
(scan plan 3)
6.4 Advanced phased array technology (scan
Scan plan 3 utilises the third leg of the ultrasonic beam from a
linear phased array probe with probe placement at skew 270. This plan 4)
examination depends on the specular reflection from the LOF. The Scan plan 4 utilises the first leg of the ultrasonic beam from a DMA
reflection signal from the LOF is highly dependent on the nature of probe with probe placement at skew 270. The DMA probe is of the
the weld cap ripples, as the ultrasonic beams get distorted and mode transmit/receive longitudinal (TRL) type, in which the refracted
converted at the ripples of the weld crown, resulting in a positional longitudinal wave is transmitted by one array and is received by
shift. another. It has the advantage of sweeping a wide range of angles
The S-scan images and observed results (amplitude, SNR and (up to 87º) with a better focusing option when compared to a linear
height) of scan plan 3 over LOF1 and LOF2 are tabulated in Tables array probe. The S-scan images and observed results (amplitude,
7 and 8, respectively. Sizing is performed using the tip diffraction SNR and height) of scan plan 4 over the LOF1 and LOF2 are
method. tabulated in Tables 9 and 10, respectively. Sizing is performed using
This examination has to be applied with caution as the additional the maximum amplitude[4] method. It is observed that scan plan 4 is
dB required to evaluate the defect and the SNR depends on the able to detect LOF1 and LOF2 with accurate positioning and a good
nature of the weld crown ripples. In this selected case study, an signal-to-noise ratio.
LOF1
Reference dB +12 dB
Practical Simulation Practical Simulation
LOF2
Reference dB +12 dB
Practical Simulation Practical Simulation
LOF1
Practical Simulation
Table 10. S-scan image and results of scan plan 4 for LOF2
LOF2
Practical Simulation
Table 11. S-scan image and results of scan plan 5 for LOF3 (left) and LOF4 (right)
LOF3 LOF4
Practical Simulation Practical Simulation
Table 12. S-scan image and results of scan plan 6 for LOF3
LOF3
Reference dB +25 dB
Practical Simulation Practical Simulation
Table 13. S-scan image and results of scan plan 6 for LOF4
LOF4
Reference dB +32 dB
Practical Simulation Practical Simulation
dB is required for good detection of the tip diffraction signals. noise ratio, which is similar to the results observed in the earlier
However, the additional dB depends on the nature of the LOF tip case study of the 12.5 mm-thick specimen.
and may vary considerably from case to case.
Table 14. S-scan image and results of scan plan 7 for LOF3
LOF3
Reference dB +18 dB
Practical Simulation Practical Simulation
Table 15. S-scan image and results of scan plan 7 for LOF4
LOF4
Reference dB +15 dB
Practical Simulation Practical Simulation
The consolidated observations for the 12.5 mm-thick and at increased dB over the reference results in the LOF tip signals
20 mm-thick weld specimens are tabulated in Tables 18 and 19, entirely merging with noise and irrelevant signals. Practically,
respectively. this condition may result in false calls or missed calls by
considering noise and irrelevant indications. Both scenarios
are possible and this may vary from one weld to another
9. Conclusion depending on the thickness and material property. Caution is
Within the scope and limited trials of this paper, a comparison recommended while using this technique, unless this is only
of the ultrasonic weld examination results of two-sided access possible option left. Additional specific training for the NDT
versus restricted single-sided access provides the following operator and validation is strongly required.
conclusions: l The conventional phased array technology using the third leg
l The conventional phased array technology using tip diffraction ultrasonic beam for examination with a weld crown is more
Table 16. S-scan image and results of scan plan 8 for LOF3
LOF3
Practical Simulation
Table 17. S-scan image and results of scan plan 8 for LOF4
LOF4
Practical Simulation
unreliable than the above as the reflected LOF response is highly – DMA refracted shear wave for complex configuration, such
distorted or completely missed due to beam interaction with as nozzles from carbon steel material;
the weld crown ripples. Practically, this possibility creates more – Detection probability for LOF existing on the same bevel
confusion for the interpreter and this option should be used (which is less responsive than the opposite side LOF).
with caution or avoided. However, this option will prove better
if the weld crown is flushed to merge with the parent material, Acknowledgements
in which case the beam reflected from the flattened weld This paper was presented at the MINDTCE-15 conference, held
crown reaches the opposite bevel face as per the scan plan. The in Kuala Lumpur, Malaysia. Input from conference delegates and
quality of machining may have an effect on achieving better peer reviewers has been incorporated into this paper to make it
results. more comprehensive. We would like to thank Mr Sirajul Hameed
l The advanced phased array technology using a dual matrix Masroor for his review and feedback.
array probe with a transmit-receive longitudinal wave (DMA-
TRL) to detect far-side bevel defects is a very efficient technique References
as, in many industrial applications, the weld being ground 1. P R Dheeraj, I Mohsin, S K Mohiuddin and S H Masroor,
flush is not desired due to economic or design needs. The LOF ‘Effect of focal law parameters on probability of detection in
responses were accurately positioned with a good signal-to- phased array ultrasonic testing using a simulation and case
noise ratio, thereby making it easy for the interpreter assessing study approach’, Materials Evaluation, Vol 74, No 11, pp 1574-
the weld. 1591.
l Many applications based on DMA probe technology can be 2. ASME Boiler Pressure Vessel Code, Section V – Non-
further studied, although this is not part of the present case Destructive Examination, ASME, July 2015
study. The following are some examples: 3. US NRC, ‘Applying ultrasonic testing in lieu of radiography for
volumetric examination of carbon steel piping’, PNNL-24232, gaps – an interim report’, April 2010.
September 2015. 10. M Moles and E Ginzel, ‘Phased arrays for small-diameter,
4. US Department of Energy, ‘Replacement for radiography thin-walled piping inspections’, 18th World Conference on
with ultrasonics for the non-destructive inspection of welds – Non-Destructive Testing, Durban, South Africa, 16-20 April
evaluation of technical gaps’, PNNL-19086, April 2010. 2012.
5. M Darmon, R Raillon, V Dorval and S Chatillon, ‘Generic 11. F Foucher, P Dubios, V Gaffard, H Godinot, H Romazzotti, A
GTD-Kirchhoff scattering model for the ultrasonic response Courbot and E Schumacher, ‘Validation of the simulation of
of planar defects’, 10th International Conference on NDE in pipeline girth welds PAUT inspection’, ASNT Fall Conference,
Relation to Structural Integrity for Nuclear and Pressurized Orlando, Florida, USA, 29 October-1 November 2012.
Components, Cannes, October 2013. 12. R Raillon, G Toullelan, M Darmon, P Calmon and S Lonne,
6. M Moles, ‘Phased arrays for general weld inspections’, 10th ‘Validation of CIVA ultrasonic simulation in canonical
ECNDT, Moscow, 2010. configurations’, 18th World Conference on Non-Destructive
7. G Grün, ‘Considerations about ultrasonic inspection of welded Testing, Durban, South Africa, 16-20 April 2012.
joints using phased array’, 5th International Conference on 13. E Ginzel and D Stewart, ‘CIVA modelling for pipeline zonal
Structural Integrity of Welded Structures, Romania, November discrimination’, NDT.net, 2011.
2007. 14. B Chapuis, F Jenson, P Calmon, G DiCrisci, J Hamilton and
8. T Armitt, ‘Phased arrays are not the answer to every application’, L Pomié, ‘Simulation-supported POD curves for automated
9th ECNDT, Berlin, 2006. ultrasonic testing of pipeline girth welds’, Welding in the World,
9. T L Moran, P Ramuhalli, A F Pardini, M T Anderson and S R Vol 58, No 4, pp 433-441, 2014.
Doctor, ‘Replacement of radiography with ultrasonics for the 15. A S Birring, ‘Ultrasonic phased array for weld testing’, Materials
non-destructive inspection of welds: Evaluation of technical Evaluation, Vol 66, No 3, pp 282-284, 2008.