Ultrasonic inspection of electron beam welded joints in · PDF fileUltrasonic inspection of electron beam welded joints in copper ... EB-weld, ultrasonic testing, inspection of copper
Post on 06-Feb-2018
229 Views
Preview:
Transcript
POSIVA OY
Working Report-97-34e
Ultrasonic inspection of electron beam
welded joints in copper
Harri Jeskanen
Pentti Kauppinen
VTT Manufacturing Technology
Materials and Structural Integrity
August 1997
M i konkatu 15 A , FIN-001 00 HELSINKI, FINLAND
Tel. +358-9-2280 30
Fax +358-9-2280 3719
Working Report-97-34e
Ultrasonic inspection of electron beam
welded joints in copper
Harri ..Jeskanen
Pentti Kauppinen
VTT Manufacturing Technology
Materials and Structural Integrity
August 1997
m MANUFACTURING TECHNOLOGY
A Work report B Public report
C Confidential report x Title of report
Ultrasonic inspection of electron beam welded joints in copper Client/sponsor of project and order
Outokumpu Poricopper Project
Kuparikapselin EB-hitsisauman ultraaanitarkastus Author(s)
Harri Jeskanen, Pentti Kauppinen Keywords
EB-weld, ultrasonic testing, inspection of copper Summary
Report No.
VALC-340 Project No.
V6SU00583 No. of pages/appendices
27 +8 appendices
The copper canisters used for the high active waste disposal are closed by electron beam welding (EB welding) and the integrity of the weld is inspected by using ultrasonic techniques. The main difficulties in the ultrasonic inspection of copper are the attenuation of the ultrasonic energy due to the large grain size and the ultrasonic noise caused by the weld. For reliable results the technique used has to be optimized and special care must be taken in the selection of ultrasonic transducers.
In this study the ultrasonic inspection techniques for inspection of EB-welds in copper canisters were developed. In the experimental work both simplified test specimen and a full scale weld specimen were used. The results verify the importance of selection of transducers when small discontinuities are to be detected in material strongly attenuating ultrasonic waves. For the inspection focussed normal incidence probes and focused angle beam probes are recommended. The focal areas of the probes should cover the complete wall thickness. Based on the experimental results it can be estimated that in EB-weld planar structural discontinuities having diameter larger than 5 mm and locating perpendicular to the sound beam can be reliably detected. With the normal incidence probe the root defects in the weld can be revealed and based on these the depth of weld penetration can be assessed.
Date Espoo _ 21.8.1997
\n_ \(___ c Rauno Rintamaa Research Manager Distribution:
Client, 3 copies
Cf~' it' ' PenttiK~n Project Manager
VTT Manufacturing TechnologyN AL6, 3 copies Authors, 2 copies
VTT Manufacturing Technology Materials and Structural Integrity P.O. Box 1704 FIN -02044 VTT, Finland
Phone: Telefax: E-mail: WWW:
~~ ?w-Checked
+358 9 4561 +358 9 456 7002, +358 9 456 5875 Pentti.Kauppinen @vtt.fi http://www.vtt.fi/manu/
Working Reports contain information on work in progress
or pending completion.
The conclusions and viewpoints presented in the report
are those of author(s) and do not necessarily coincide
with those of Posiva.
2
Contents
Tiivistelma 3
Abstract 4
1. Introduction 5
2. Test programme and the specimens 5
3. Sun1ffiary of literature survey 6
4. Experimental ultrasonic techniques 6 4.1. Mode conversion technique
4.1.1 Creeping wave on the scanning surface 6 4.1.2 Creeping wave on the surface opposite to 7
scanning surface 4.2. Shear wave probes 8
4.2.1 Tandem-technique 8 4.3. Longitudinal wave probes 8
5. Results of experimental measurements 8 5.1. Inspection of welds
5.1.1. Normal incidence probe Panametrics V104 10 5.1.2 Scanning with angle beam probe 10
RTD70°TRL2-Cu 15x22)SA15°FD 5.1.3 Scanning with angle beam probe 11
RTD60°TRL2 Cu2(15x22) SA3°FD 5.2 Reference defects
5.2.1. The normal incidence probe Panametrics V 104
5.2.2 Scanning with angle beam probe RTD70°TRL2-Cu2(15x21)SA15°FD
5.2.3 Scanning with angle beam probe RTD60°TRL2-Cu2(15x22)SA3°FD
5.3. Inspection of the full scale weld sample
11
12
13
14
6. Summary and conclusions 15
7. References 17
~----------------···---
3
Tiivistelma
Y dinpolttoaineen loppusijoitukseen kaytettavat kuparikapselit suljetaan
elektronisuihkuhitsauksella (EB) ja liitoksen eheys varmistetaan ultraaanitestauksella.
Kuparin ultraaanitarkastusta vaikeuttaa materiaalin suuri raekoko ja hitsin
ultraaanisignaaliin aiheuttama kohina. Luotettavan tarkastustuloksen saavuttamiseksi
tarkastustekniikka on optimoitava ja erityisesti on kiinnitettava huomiota kaytettavien
ultraaaniantureiden valintaan.
Tassa tutkimuksessa kehitettiin EB-hitsin tarkastustekniikkaa kaytUien seka
yksinkertaistettuja koekappaleita etta halkaisijaltaan taysimittaista kuparikapselia.
Koetulokset osoittavat anturien valinnan tarkeyden pyrittaessa havaitsemaan
luotettavasti pienia epajatkuvuuskohtia voimakkaasti ultraaanta vaimentavassa
materiaalissa. Tarkastukseen suositellaan kaytettavan seka fokusoituja
normaaliluotaimia etta fokusoituja kulmaluotaimia, joilla katetaan koko tutkittava
paksuusalue. Koetulosten perusteella EB-hitsista havaitaan halkaisijaltaan 5 mm
ylittavat tasomaiset heijastajat, jotka sijaitsevat kohtisuorassa aanikeilaa vastaan.
N ormaaliluotaimella havaitaan hitsin juuriviat, joiden perusteella voidaan arvioida my os
hitsin tunkeuman syvyytta.
4
Abstract
The copper canisters used for the high active waste disposal are closed by electron beam
welding (EB welding) and the integrity of the weld is inspected by using ultrasonic
techniques. The main difficulties in the ultrasonic inspection of copper are the
attenuation of the ultrasonic energy due to the large grain size and the ultrasonic noise
caused by the weld. For reliable results the technique used has to be optimized and
special care must be taken in the selection of ultrasonic transducers.
In this study the ultrasonic inspection techniques for inspection of EB-welds in copper
canisters were developed. In the experimental work both simplified test specimen and a
full scale weld specimen were used. The results verify the importance of selection of
transducers when small discontinuities are to be detected in material strongly attenuating
ultrasonic waves. For the inspection focussed normal incidence probes and focused
angle beam probes are recommended. The focal areas of the probes should cover the
complete wall thickness. Based on the experimental results it can be estimated that in
EB-weld planar structural discontinuities having diameter larger than 5 mm and
locating perpendicular to the sound beam can be reliably detected. With the normal
incidence probe the root defects in the weld can be revealed and based on these the
depth of weld penetration can be assessed.
l
5
1. Introduction
Copper canisters will be used for the high active waste disposal of the Finnish nuclear
power plants. The filled canisters are closed by welding a copper lid to the canister. For
welding a narrow gap welding technique, electron beam (EB) welding has been
proposed. The joint between the canister and the lid has to be inspected after welding in
order to be sure that no significant defects are existing in the weld. There ate two
possible inspection techniques for the welds in thick copper plates: ultrasonic
inspection and radiography /11. The ultrasonic testing of copper is more difficult than
the conventional testing of steel and similarily to steel the weld joints are making the
inspection even more difficult.
The ain1 of the present work was to study the applicability of ultrasonic techniques for
inspection of EB-welds in copper canisters. The practical measurements were performed
with several specimen simulating the real weld geometry in the copper canister. The first
simplified specimen containing artificial reflectors were used to test different ultrasonic
techniques and to optimize the transducers used in later experiments. The final tests
were performed with a specimen representing in full size the EB-weld between the lid
and canister. The results of the experimental measurements and the conclusions from
the work performed are presented in this report.
2. Test programme and the specimens
The ultrasonic measurements were performed using three different test specimens. The
first specimen was a hot rolled copper plate containing an EB-weld. The specimen is
shown in appendix 1.
The other two specimens are shown in appendix 4. The weld geometry simulates the
geometry of real copper canister but the specimen is not curved as the real canister
surface.
6
3. Summary of literature survey
A literature survey was performed in order to find usefull hints for practical
measurements. The result of the survey was, however, that only very limited amount of
data has been published about ultrasonic inspection of welds in thick copper products.
The most use full information was given in the report "A study of attenuation and
scattering of ultrasound in polycrystalline copper" published by Swedish Nuclear
Power Inspectorate (SKI) in Sweden /2/. This report summarizes the major difficulties
1net when inspecting copper material. Based on the report one of the main difficulties,
the grain size, is caused by the fact that it is difficult to achieve fine grain sizes in thick
copper forgings. The structure of a rolled thick plate is a mixture of many fine grains
and relatively few coarse or very coarse grains. The recrystallised structures of hot rolled
plate are heavily twinned and these twins can be very effective reflectors of ultrasound.
If suitably oriented the twins in large grains may increase the noise to the ultrasonic
signal.
4. Experimental ultrasonic techniques
4.1 Mode conversion technique
4.1.1 Creeping wave on the scanning surface
The creeping wave propagating along the scanning surface can be used for detection of
surface opening defects and the simultaneously created direct longitudinal wave having
large angle of incidence for detection of defects just below the surface (figure 1).
In the experimental measurements both single crystal and twin crystal probes were used.
The twin crystal probe focused at the depth of 10 mm below the surface and having
separate crystals for transmitting and receiving ultrasonic pulses (SE probe) proved to be
more effective than the single crystal probe.
7
Direct longitudinal ~ • .. ":..J.---1'---t-----1---/
30-70-70-wave (K1) ............ . ... >...... I
................... , I
..... 30-70-70-30-wave (~:i Direct shear wave
Creeping wave ID
Figure 1. The principle of mode conversion. The creeping wave propagating along the
surface and the direct longitudinal wave can be used for detection of surface opening
defects and defects close to the scanning surface.
4.1.2 Creeping wave on the surface opposite to scanning surface
The creeping wave propagating along the surface opposite to scanning surface can be
used for detection of incomplete weld penetration. The application of this technique is
restricted only to the base metal of the hotrolled plate because from the interface
between the weld and base metal a metallurgical indication will be recorded due to the
large grain size in the weld material. Furthermore, the large grain size causes strong
attenuation of ultrasonic waves especially in the case of shear wave probes.
The mode conversion 30-70-70 (26-70-70) can be used for rough estimation of flaw
size. If the incident angle of shear wave is 26° the angle for longitudinal wave is 70°.
4.2 Shear wave probes
Due to the strong attenuation of shear waves in the base metal the use of shear wave
probes is impossible in practise. From the different shear wave probes experimentally
8
tested the best results were achieved with a composite probe. When measuring the
reference holes in the weld of specimen 1 the signal-to-noise (SIN) ratio was 3-5 dB
with the 2 MHz probe (SWB702).
The use of large incident angles on a perspex-copper interface is for physical reasons not
possible. If large incident angles are necessary the inspection must be performed in . . Immersion.
4.2.1 Tandem-technique
In principle, the tandem-technique would be optimal for detection of planar defects in
the current weld geometry. However, the use of this technique is restricted by the strong
attenuation of shear waves.
4.3 Longitudinal wave probes
In the ultrasonic inspection of copper the longitudinal wave probes are clearly most
effective. The attenuation of longitudinal waves in copper is remarkably lower than the
attenuation of shear waves having same nominal frequency. With shear waves low
frequencies have to be used to overcome the problems caused by attenuation.
5. Results of experimental measurements
In the measurements the automatic ultrasonic inspection system Sumiad Ill has been
used. The logarithmic amplifier used in the measurements allows the measurement of
echo amplitudes having differences up to 80 dB. In the record sheet V65-6353-1
attached in appendix 8 the principle of inspection and the scanning parameters are
presented.
~----------------------- - ---
9
The defect types detected are root defects in the weld, incomplete penetration and
"craters" on the outside surface. These craters are both surface opening and close to the
surface. In addition, bonding defects (incomplete fusion, lack of fusion) can be expected
when the position of EB-beam is not precisely in the centre of the weld groove.
Specimen 1 was inspected with several shear and longitudinal wave probes that are
normally used for inspection of steel. Based on the results of these measurements the
technique described below was specified and the probes were optimized to inspection of
copper.
The scanning with normal incidence probe is made from the forged side of the specimen
as shown in figure 9. The grain size in the weld and forged base metal limits the
frequency of probe that can be used. In order to have optin1al resolution the diameter of
the sound beam in the weld area should be as small as possible. The probe used in these
measurements was not quite optimal. Defects close to the surface or opening to the
surface are not detected in scanning with normal probe due to the deflection of the
sound beam.
When the scanning is performed from the hot rolled side of the specimen the defects
close to surface or opening to the surface can be detected with creeping wave probe
emitting also longitudinal waves (figure 10).
Both plates I-32 and I-20 have been tested first before cutting the specimen in two
pieces as shown in appendix 4. The reference reflectors machined in the specimen are
shown in appendices 2 and 3.
The cut specimen I-32-1 and I-32-2 have been inspected with acoustic microscope
before machining the reference reflectors. In this inspection a focussed 5 MHz probe
having crystal diameter of 2 inches has been used. The scanning was performed in
immersion from the surfaces shown in appendix 4. The resolution of inspection
performed in immersion is essentially higher due to the higher frequency used and the
shorter sound path in copper. The result of this inspection can be seen in figure 8.
10
5.1 Inspection of welds
5 .1.1 Normal incidence probe Panametrics V 104
Figures 2 and 3 show the results of inspection of the welds in specimens I-20 and I-32.
The welding defects in the root area can be clearly seen in the images. It can be also
noticed that that the weld penetration is increasing in the direction of welding.
In specimen I-32 the noise level of the weld is higher (both images are measured at same
level). The difference in noise level is caused by different welding parameters; in this
case the lower voltage used in welding has increased the width of weld.
Results of inspections:
•
•
•
In specimen I-32 the SIN-ratio in weld area is 3-13 dB compared to a 0 3
mm cylindrical hole.
In specimen I-20 the SIN-ratio in weld area is 6-18 dB compared to a 0 3
mm cylindrical hole.
Welding defects were not detected in these specimens .
The weld penetration in specimen I-32 was 46-51 mm and in specimen I-20 46-55 mm.
The depth of weld penetration has been measured by dropping the echo amplitude from
the root defect to 50 o/o from its maximum value.
5.1.2 Scanning with angle beam probe RTD70°TRL2-Cu (15x22)SA15°FD-10
In the inspections performed from the hot rolled sides of the specimens primarily
surface opening defects were detected. The defects were short in X -direction and
therefore the maximum shift of the focussed probe between scannings in X -direction is
3 rrun. The surface contour was not optimal to the relatively large probe. The surface
11
quality is an important factor affecting the reliability of inspection and has therefore to
be taken into account in future measurements. Both the surface roughness and the
possible waviness of the surface should be as low as possible to avoid disturbances in
the acoustical coupling between the probe and the material. The requirement for surface
roughness is Ra < 6,3 J.lm. The acceptable waviness is depending on the length of the
contact surface of the probe used in the inspection. In practise unacceptable waviness is
caused by grinding the surface manually.
The figures 4 and 5 show the results of specimens I-20 and I-32.
5.1.3 Scanning with angle beam probe RTD60°TRL2-Cu2(15x22) SA3°FD-40
With the 60° angle beam probe the planar defects can' t be detected. The probe that is
focused at the depth of 40 mm is used for inspection of defects in the weld root area and
to measure the depth of weld penetration.
Figures 6 and 7 show the results of measurement of plates I-20 and I-32 with the 60°
probe.
5.2 Reference defects
5.2.1 The normal incidence probe Panametrics V104, 2.25 MHz, 0 l "
The plate I-32-1 has been tested with the normal probe from the forged side of the joint
as shown in the figure 9. The cylindrical holes both in the front of weld and behind the
weld are clearly detected. The SIN-ratio in the weld area is 7-18 dB compared to the
cylindrical holes.
Most of the flat bottom holes are also detected. The lowest SIN-ratio is 3dB below the
maximum noise when compared to the maximum noise level of the weld. This means in
practise that defects possibly existing in the area of maximum noise level in the weld
would not be detected.
For reliable detection of a flaw the echo amplitude measured from the flaw should
exceed the maximum noise level by at least 6 dB. In the weld of specimen 1-32-1 the
12
echo amplitude measured from the flat bottom holes is 2 - 5 dB higher than the
maximum noise level of the weld. From the formula d2 =(HI I H2) d1, where di = 3mm
and the ratio between the heights of echos HI and H2 is (HI I H2 ) = 1.1 - 1.6, the
minimum diameter of the detectable flat bottom hole is d2 = 3.3 - 4.8 mm. On the base
metal side of the joint the minimum diameter of detectable flat bottom hole is 3 mm.
The SIN -ratio of this hole is more than 6 dB both in front of the weld and behind the
weld. The flat bottom holes locating close to the surface are not detected due to the
deflection of the sound beam and due to the adverse effect of the edge of the specimen.
Table 1. Plate 1-32-1; scanning with normal incident probe. The echo amplitudes
measured froJn reference holes shown in figure 9 have been compared to the echo
amplitude from the cylindrical hole C (depth 55 mm, diameter 0 3 mm)
Cylindrical Echo Flat Echo Flat Echo Noise from
hole amplitude bottom amplitude bottom amplitude the weld dB
dB hole dB hole dB
A -1 E2 -12 T2 -10 -7- -18
B -1 E3 -10 T3 -9
c 0 E4 -9 T4 -10
D 0 ES -9
5.2.2 Scanning with angle beam probe RTD70°TRL2-Cu2(15x21)SA15°FD-10
Noise
from
the
base
metal
dB
-16--
20
The specimen I-32-2 was tested with a longitudinal wave angle beam probe as shown in
figure 10. The probe was focused approximately at the depth of 10 mm and thus it can
be used for detection of surface opening defects by the creeping wave component. In
13
addition the longitudinal wave component can be used for detection of defects to the
depth 15 mm below the surface.
As presented in table 2 the 03 mm flat bottom hole is detected with SIN-ratio of 2 -9
dB. Correspondingly the cylindrical holes are detected with SIN-ratio of 13 - 23 dB. The
attenuation of ultrasonic in the weld is not remarkable (compared with measurements
performed with cylindrical holes). The difference between results measured in front and
behind the weld is probably mainly caused by differencies in the acoustic coupling of
the probe.
Table 2 Plate 1-32-2; angle beam probe 70 ° SEL. The echo amplitudes of reference
holes shown in figure 10 compared to the echo amplitude from the cylindrical hole A15
(on the base metal side of the joint, distance to surface 15 mm, 0 3 mm)
Cylindrical Echo Flat Echo Notch Echo Noise from Noise from
hole amplitude bottom amplitude depth mm amplitude the weld the base
dB hole dB dB dB metal dB
AS -2 El -17 Ul -19 -23- -30 -30- -33
AlS 0 E2 -17 U2 -16
BS -10 Kl -14
BlS -2 K2 -21
The noise caused by the weld has been measured at the depth of 0 -15 mm and the noise
from the base metal on the hot rolled side of the joint.
5.2.3 Scanning with angle beam probe RTD60°TRL2-Cu2(15x22)SA3°FD-40
The specimen I-32-2 was tested with the 60° longitudinal wave probe as shown in figure
10. The probe is focussed at the depth of 40 mm as can be seen from the results shown
in table 3. As expected planar defects are not detected with this probe. The noise from
the weld is not remarkable. This indicates that the grain structure of the weld has certain
orientation (compare the results with oo and 70°probes).
14
Table 3 Plate I-32-2; Angle beam probe 60° SEL. The markings of cylindrical holes
refer to figure 10. The number in the identification shows the depth location of the
cylindrical hole.
Cylindrical Echo amplitude Cylindrical Echo amplitude Noise from the Noise from the
hole dB hole dB weld dB base metal dB
identification
AS -15 B5 -21 - -25 -21 - -25
A15 -6 B15 -10
A25 -2 B25 -6
A35 -1 B35 -5
A45 0 B45 -4
5.3 Inspection of the full scale weld sample
The ultrasonic inspection of the full scale specimen was performed 25.-26.4.1997 in the
facilities of Outokumpu Poricopper Oy in Pori. The test arrangement is shown in the
photographs. The testing was performed by using the technique and transducers
described in 5 .1. In the position where welding process has been started and stopped a
defect having a length of approximatelly 250 mm was detected. It was assessed that
this defect is formed by several individual small defects. The defect is starting from the
inside surface and is propagating gradually to the outside surface thus extending as a
"cloud" through the whole wallthickness. In addition, several defects (crater type) were
detected in areas close to the surface and even at some depth inside the material. The
results of ultrasonic testing with different transducers are shown in figures 11 -16. From
the scannings performed with angle beam probes only the results from scannings on the
side of the hot -rolled plate are presented.
15
6. Summary and conclusions
This report describes possible techniques to solve the inspection problem of the EB
welded copper canister. It is, however, possible that even more suitable techniques could
be developed. The most serious problem in the ultrasonic inspection of the EB-weld in
copper is the strong attenuation of ultrasonic waves in the weld material. The
attenuation is also increased due to the large grain size of the base metal. One solution
to overcome this type of problem would be to use signal processing techniques.
However, the use of e.g. SAFf -algorithm to process the complete measurement data is
not yet possible with current computers. Nevertheless, the main problem is to transmit
sufficient ultrasonic energy into the material and this can be only solved by using a
probe having large piezoelectric crystal.
Also the "kissing bond" -effect where the surfaces are in close contact but there is not
real bonding between the surfaces can cause remarkable problems in inspection.
Only when the problems described here are solved (or the adverse effects minimized)
the reliability of inspection and the minimum size of the detectable defect can approach
the level that is normally achieved in the ultrasonic inspection of steel. There are a few
methods to minimize the adverse effects of weld in inspection:
•
•
•
The ultrasonic noise caused by the weld is decreased when the angle
between the ultrasonic beam and the weld/base metal interface is larger
than 0°.
The transducer used can still be optimized. The normal incidence probe
has to be focussed and the applicability of a twin-crystal transmitter
receiver (SE) probe should be tested. With the technique used in this
study the volumetric defects are reliably detected but the technique for
detection of planar defects needs improvement.
The normal incidence probe can be improved by following means:
•
16
A transmitter-receiver probe with nominal frequency 2.25 MHz will be
constructed. The piezoelectric element of this probe is formed from
composite crystals having such diameter that the focus of the probe is in
the depth zone to be studied. The angle between the crystals and the
normal of the surface is 10 - 20°. With this type of transducer the noise
caused by the weld will most probably be reduced and smaller defects
could be detected.
Scanning could be performed with several angle beam probes being
focussed at different depths to cover the whole wallthickness.
With the technique described above the planar reflectors having diameter
more than 5 mm and corresponding to a flat bottom hole can be detected
in the EB-weld when the reflectors are located perpendicularly to the
incident bemn (material as in sample I-32/I-20 and grain size 120 J.lm or
less). If the grain size is larger the size of the defect that can be detected is
also remarkably growing.
In the scanning of the full scale specimen performed with normal incidence probes not
even the large craters detected by angle beam probes could be reliably seen. Based on
this it could be assessed that the craters are formed from several smaller reflectors
having diameters less than 5 mm.
The root defects in the weld can be clearly detected with normal incidence probe in the
weld of the full scale specimen. The depth of weld penetration can be evaluated from
the root defect. Also the possible defects caused by wrong positioning of the electron
beam in welding can be detected either as a change of echo amplitude (in the case of a
large reflector) or as a shift of noise area in the cross sectional image (D-scan image). In
angle beam scanning the craters are detected at different depths due to their 3-
dimensional nature.
17
7. References
/1/ Aalto, H., Rajainmaki, H., Laakso, L. 1996. Production methods and costs of
oxygen free copper canister for nuclear waste disposal. Report Posiva 96-08, Posiva Oy,
Helsinki.
/2/ Bowyer, W.H. and Crocker, R.L. 1996. A study of attenuation and scattering of
ultrasound in polycrystallinen copper. SKI Report 96:27. Swedish Nuclear Power
Inspectorate, Stockholm. 93 p.
18
Figure 2 The joint 1-20 has been inspected by scanning with a normal incidence probe (0°)from the forged side of the joint. Welding directionfrom right to left. Weld penetration 45-65 mm. No remarkable welding defects except root defect in the weld can be seen in the images.
Figure 3 The joint 1-32 has been inspected by scanning with a normal incidence probe (0°)from the forged side of the joint. Welding direction from right to left. Weld penetration 46-51 mm. The noise coming from the weld is higher than in figure 2 due to the fact that the width of the weld is larger.
19
Figure 4 The echo amplitudes of indications detected in joint /-20 are -17- -21 dB when compared to the echo from a diameter 3 mm cylindrical hole at the depth of 10 mm. The scanning of joint has been peiformed from the side of the hotrolled plate with a 70 <YJ'RL-Cu transducer.
Figure 5 The echo amplitudes of indications detected in joint /-32 are -19- -22 dB compared to the echo from a diameter 3 mm cylindrical hole at the depth of 10 mm. The scanning of joint has been peiformed from the side of the hotrolled plate with a 70 <YJ'RL-Cu transducer.
20
Figure 6 The scanning of joint 1-20 has been performed with a 60° angle probe from the side of the hotrolled plate. The blow-outs in the root area are clearly seen in the image. The weld penetration seems to be approximatelly 40 mm. This value is probably not correct because the focus sing of the probe to 40 mm causes a shift upwards in the location of the indication.
Figure 7 The scanning of joint 1-32 has been performed with a 60° angle beam probe from the side of the hotrolled plate.
21
Figure 8 Plates 1-32-1 and 1-32-2 were measured also by using a C-mode scanning acoustic microscope. In the upper image a typical "crater" on the surface can be seen. The width of this crater is 3 mm and height 10 mm. The same defect has been detected also in the inspection performed with a 70 o angle probe (figure 5 ).
The scanning of plate 1-32-1 has been performed from the side of the hotrolled plate (grain size 0.120 mm) and scanning of plate I-32-2from the forged side (grain size 0.500 ... 1.500 mm). The increase of grain size clearly causes the increase of the smallest detectable defect size. The scanning directions are presented in appendix 4.
FBHcp3mm SDHcp3mm Width of the notch - 1 mm
22
I()
<0 C\1
""
Figure 9 The ultrasonic image of joint 1-32-1. The inspection has been performed with a normal incidence probe. The upper image is the A -scan (time-amplitude) presentation, in the middle the C-scan image (top-view) and the lowest is the D-scan image (sideview). The surfaces used in scanning and scanning directions are shown in the drawing below the images. This drawing is not presenting the sample according to the projection rules of technical drawings. The refelectors shown in red colour in the drawing have been detected in ultrasonic testing and corresponding indications are marked in the images. On the side of the base metal the smallest detectable reflector is a flat bottom hole having the diameter of 3 mm. In the weld area the smallest detectable defect is flat bottom hole having diameter of 5 mm due to the strong noise caused by the weld. The uppermost (El and Tl) holes are not detected due to the adverse effect of the edge of sample and the deflection of the sound beam.
FBH$3mm SDHcp3mm Width of the notch - 1 mm
23
Figure 10 The ultrasonic image of joint /-32-2 inspected with a 70° angle beam probe focus sed at the depth of 10 mm. The direction of scanning is shown in the drawing below the images. The drawing presents the sample in the same way as drawing 9. The reflectors shown in red colour in the drawing have been detected in ultrasonic testing and corresponding indications are marked in the images. Marking "D" is used for welding defects. The acoustical noise from the weld is lower than with normal probe (figure 9 ). The variation of echo amplitudes is mainly caused by differencies in acoustic coupling. The smallest detectable defect with the 70° angle beam probe is less than 0 5 mm in the weld area.
24
Figure 11. The ultrasonic image of the weld in the full scale specimen. Measurement with normal incidence probe. In the position X=3000 mm, Z=57 mm a 3 mm cylindrical hole is located. The start point of welding is at X-coordinate 276 mm the length of starting area is 102 mm. The slope out area (ending area) is at 378-642 mm. The weld penetration is 50-58 mm.
Figure 12. Magnified image from the defect seen in figure 11 at the end point of welding.
25
Figure 13. Ultrasonic image measured with a 70°SEL-probe in X-position 0-1531 mm. The defect at the end point of welding is in position X=520- 606 mm. Also crater type defects can be seen in the image.
Figure 14. Ultrasonic image measured with a 70°SEL-probe in X-position 1531 - 3100 mm. The defects seen in the image are craters from which most are opening to the surface.
26
Figure 15. Ultrasonic image measured with a 60°SEL-probe in X-position 0-1531 mm. The defect at the end point of welding is in position X=410- 600 mm. Also crater type defects and the longitudinal weld of the cylinder (in position X=1200 mm) are seen in the image.
Figure 16. Ultrasonic image measured with a 60°SEL-probe in X-position 1531 - 3100 mm .. Crater type defects and the longitudinal weld of the cylinder (in position X=2750 mm) are seen in the image.
Appendices
Appendix 1, Plate 1
Appendix 2, Plate I 32-1
Appendix 3, Plate I 32-2
27
Appendix 4, Plate I 32 (similar to I 20)
Appendix 5, Full scale sample
Appendix 6, Macro graphs of specimen I 20
Appendix 7, Fotographs showing the testing arrangement
Appendix 8, Data sheet V 65-6353-1 Examination summary
• ~0 ~ - ~
l() <.0
I~ T=SO ~I
• •
~: ::::: ::::::::& : :7:7::J::;:~ZI~::::: : :::::::f:::'::: ::::: ::::::::::: ::: : :::::::: : : ::::::::::T~:~;:::;:r:?::::: :::::::: ::{:::::::::: : : :::::::: :::::::::::T~T'?TS~ • • - -EfS·t:Zf:~:~
1
l() -.::t C\J
•
... 3/4T ...
~ ~ tn s ~
~
Osa I Nimitys I Piir. n:o I Aines I Kpl I Huom!
Kok.piir. __ Pinta I Projisointi Pvm. Nimi
T_o~_~_r:_~~-s_it _____ d -~ 13.2 .. 97 _l!:!_ Pa1no c:J ~7}
~--- -- ------·-· ~----m PLATE 1 HJ972.dsf Valmistustekniikka
0 CO
0 N
l{)
~
m lN
~ 14 --~~-8-
L ~ I o
210 _ __j
1.0 <.0 , N
,.......
Os a
55
FBH ~3 mm SOH ~3 mm Width of the notch -1 mm
Nimitys Kok.piir. Pinta
Toleranssit Paino
itrr
Huoml
~ ~ tT:1
~ ->< N
Pvm. INimi Piir. 18.11 .96 I TH Tark.
r~
80
90
110
120
... J
l{') C\J
210
0 C\J
UJ IN ~~
0 C\J
FBH ~3 mm SOH ~3 mm
1 r
Width of the notch -1 mm
Osa I Nimitys
Kok.piir. 1 Pinta
Paino
5
--· ---w-- I CALIBRATION BLOCK I 32-2 Valmistustekniikka
Huom!
>~ ~ tJ'j
~ ~
~ VJ
Pvm. INimi 8.11.96 I TH
HJ966.dsf
0 CX)
210 ,~
CD
~~~ 210
Direction of insonification
~--------------------- -
! ----------- , - -----1---,-
0
~:.s::~~i:.::~·:·:-:-:·:·: :-:·: .·.·.·.-:-:-:-:-:·.·. :-:·:-:-: · :-:·:s.:2:.s:.:::.:l;::::.:;:;.:;_::::::~{:~:: : ::)]:~:~i~?l~z:::::::~~~;:l:::~L:i:L<7i:~:: : :: :~~;::::::~~:{2IZ~ ' N T •
Direction of insonification
® l() 1'-
- - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - , ______ x._
~ cl
Toleranssit
Paino
c · · ··.·.·.·.· . . ·.· .·.· ... .
v r r
Tark.
?; ~ t1i s ~ ~
-rn I SCHEME FOR CUTTING OF Valmistustekniikka CALIBRATION BLOCKS HJ964.dsf
.,------
l{)~ I l{)~ l{) l{) 1'-- !'--()) CO -e- -e-
N '<j" 1'--e-
F 25~15) ~I ~~~ 65
_<?.~~ .L _ ___ Nim~s __ _ Kok.piir. j Pinta
Toleranssit Paino
-m FULL SCALE SAMPLE Valmistustekniikka
Huoml
~ ~ t"I1
8 ~ Vl
Pvm. INimi Piir. 122.4.97 I TH Tark.
HJ973.dsf
APPENDIX 6
Grain size 120 lJ,m Grain size 500-1500 lJ,m
I Hot rolled EB-weld Forged
Hot rolled Weld Forged
Macro graphs showing the grain structure of sample 120 ( x 2 0)
APPENDIX 7
The arrangement used in ultrasonic testing of the full sca]e smnp1e.
APPENDIX 8
iiTT . ~
TESTAUKSEN YHTEENVETO V65-6353-1
VALMISTUSTEKNIIKKA Examination summary Sivu/Sheet
Manufacturing Technology 1 (1) T ilaaja/Co n tractor Testausohje/P rocedure
Outokumpu Copper -I UT -laitteisto/equipment
Sumiad Ill v.3.34 Testauspaikka/Test carried out at Testauskohde/Test item
Otaniemi/Pori EB-welds 1-32 , 1-20 and full scale weld
S ka n ne ri!M a-nip u la tor Y -askel/ Y -step
1.0 mm X-askel/overlappin g
3.0 mm NopeusNelocity mm/s
ETS-85
Kalibrointitiedosto/Calibration data file PAN225
Kanava Luotain
Channel Probe
1 V104
2
3
4
Luota u ssu u n n itelm a/Pian n ing
Tiedosto/Data file
Kulma Kalibrointi ptk Angle Calib. report
0 6353-2
KAPSELI .PLN
I I ! I
I I
11 u
I I
I l L.S'JJ
"" I
Y-140
~;~·:- 1
I _.__y
l-J~ X-overlapping
~------------------
I
Pvm /Date
i) '-1. ~ 7-I Poyt~n tarkast~nut/Report inspected
/ ; ,U/1£.-5--~~ Pentti Ukkonen
50
.UT
Kanava Lu otain Kulma
Channel Prob e Angle
5 96-727 70
6 96-728 60
7
8
Luotainpidike/Probe holder
Tiedosto/Data file KAPSELI
Y-90
I l
! I
i h7.,0. 6
T~ /:
A I =,.._ __ _..jj
~ "'\1 I ~/60o : I
i I
I I
I
I ' ___ __j
Kalibrointi ptk
Calib . report
6353-3
6553-4
.MOD
Pvm/Date I Testaaja/~x.am~i ~ ~ I Patevyys/Competence
:2~t1,._,_0=--'j.......L."--. ,.:...._. 9=-q.:.-<-.' )_._____ (72) ~4· . ~-( CVL/
~ PL 1714, 02044 VTT
Puh .vaihde (90) 4561
Harri Jes anen
0 PL 17031 (Kan slerinkatu 8H) , 33101 Tampere
Puh.vaihde (931) 163 111
EN 3
top related