-
17th World Conference on Nondestructive Testing, 25-28 Oct 2008,
Shanghai, China
ULTRASONIC TESTING OF AUSTENITIC WELDS AT THE RUSSIAN NPP
Vladimir G. BADALIAN, Аlexey Kh. VOPILKIN, Pavel F. SAMARIN,
Dmitry S.
TIKHONOV, SPC ECHO+, Moscow, Russia
E-mail: [email protected], [email protected],
[email protected], [email protected],
Web: http://www.echoplus.ru
Abstract
Since 1997 Scientific and Production Center ECHO+ is maintaining
automated ultrasonic testing of austenitic welds Du300 at the
Russian NPP with RBMK reactors. Testing is executing with AUGUR
series systems. The main feature of AUGUR system is a coherent data
treatment and measurement of real flaw size. On the NPP units under
exploitation were inspected more then 10 000 welded joints Du300.
Comparison between flaws sizing by AUGUR methodology and by
metallographic investigation was completed and absolute height
sizing error was less than ±2 mm in 95% confidence interval. UT
with coherent treatment allowed to proceed monitoring of flaws and
to make planning of NPP maintenance more flexible. With a new
generation system AUGUR 5.2 since 2004 is maintaining testing of
welds on steam generators sockets: composite socket weld Du1100 of
VVER 440 steam generator and perlitic socket weld Du1200 of VVER
1000 steam generator. In these welds were detected intolerable
flaws and comparison between UT and metallographic data shown good
agreement. Recently inspection procedures for a various austenitic
welds were developed: Du200 pipeline welds, Du500 primary circuit
pipeline longitudinal and circular welds, primary circuit pipeline
and main shutoff valve socket weld, ring of main circuit pump
socket weld, welded joints of turbine house, welded joints of
pressure compensator. All procedures were validated on mock-ups
with real and realistic flaws. This paper presents main stages of
inspection procedures validation, peculiarities of UT with AUGUR
5.2 system and examples of flaws sizing. Keywords: ultrasonic
testing, coherent treatment, FT-SAFT, austenitic welds
Introduction
The problems related with austenitic weld with ultrasonic
methods (UT) well-known [1]. Solving of this problem is one of the
aims of Scientific and Production Center ECHO+. The main task is
increasing of signal-to-noise ratio (SNR) – the main factor of
flaws detectability and sizing success. Were tried some methods:
space filtration and superresolution methods [2], spectrum
splitting method with further spectrum extrapolation [3], two-mode
and two-frequency methods [4], coherent data treatment with FT-SAFT
[5], image reconstruction with use of wave transformation on the
flaws. All of these methods were implemented with AUGUR 5.2 system;
some of them were included and accepted in inspection procedures of
austenitic weld on Russian NPP. This paper includes description of
this procedures and hardware acceptation and implementation for
operating inspection.
Base austenitic welds inspection procedure
The prevalent technique of UT testing is longitudinal wave
pitch-and-catch technique with probes focused at particular wall
thickness [6]. Inspection procedures for AUGUR systems includes
some techniques to increase SNR apart from the selection of optimal
sounding scheme and frequency diapason, accounting specific weld
peculiarities [4]:
• two-mode inspection with shear and longitudinal waves
simultaneously;
mailto:[email protected]:[email protected]
-
2
• space spectrum filtering of source B-scans; • coherent data
treatment; • image reconstruction with account of wave mode
transformations of the flaws.
The probes used for two-mode inspection has similar directivity
patterns for longitudinal and shear waves. The beam divergence is
rather high to increase the angles of incidence range for further
combined coherent treatment. Generally probes with two or three
various main angles of incidence used so for example for 55 mm wall
thickness weld the range of incidence angles is about 70° and more
taking into account the shear wave emitted by longitudinal wave
plate.
Is search mode all possible techniques used to increase
detectability. SNR increase with use of longitudinal and shear wave
in pulse-echo mode is reached with difference of various flaws
reflection power. E.g. some plane flaws is better detected by shear
waves and volumetric flaws by longitudinal waves. This difference
accentuates with coherent treatment which could increase the SNR at
4-5 dB for shear waves and 6-10 for longitudinal waves [7, 8].
Coherent treatment is increasing the amplitude of indications which
correspond to reflection from flaws and decreases the contribution
of multiple scattering, reverberation and other noise indications.
Henceforth the base inspection procedure combines effective methods
of structural noise suppression which is important task is
austenitic weld testing. Principal concern was taken into account
for image analysis methodology.
Austenitic welds with wall thickness < 19 mm inspection
procedure
For adaptation of base procedure to test circular austenitic
welds with diameter 325 mm (Du300) and thickness from 10.5 to 10 mm
were completed experiments and numerical calculation to accept
probes parameters for optimal sensibility and resolving power. Were
suggested to combine two plates in one probe case: shear wave plate
with 4 MHz and longitudinal wave plate with 5 MHz. The using of
these probes allowed to obtain high resolving power, increase SNR
with use of two-mode inspection.
Figure 1. UT and metallographic data comparison
To determine sizing accuracy of longitudinal flaws trials on the
artificial (notches) and
real flaws (corrosion cracks) were realized. Comparison between
UT datasheets and mock-ups datasheets and results of real cracks
metallographic investigations was completed. Maximal length sizing
accuracy in 95% confidence interval is 6 mm for notches. Maximal
height sizing accuracy in 95% confidence interval is 1 mm for
notches.
a) Flaw image b) Photo of the templates
-
3
Figure 2. Absolute height sizing accuracy histogram (Du300)
At the Figure 1 shown the crack B-image obtained with shear
waves and photos of
templates with this crack. Dark color of scurf allows to see the
crack profile. At the Figure 2 shown the histogram of UT and
metallographic height measurement for eleven cracks. For 95% (62
from 65) measurements the accuracy is between -1.5 mm and 1.5 mm.
The procedure attestation was allowed to accept it for operating
inspection on the NPP with RBMK reactors and expand it to same type
welds. Austenitic and composite welds with wall thickness > 20
mm inspection procedure
Described approach was used in task of development inspection
procedure for thick wall austenitic and composite welds at VVER-440
and VVER-1000 reactors. On trials were used real welds with
artificial and natural flaws. As the result of trials were accepted
new inspection procedures with use of AUGUR 5 system:
• primary circuit pipeline (Du500) longitudinal and circular
welds;
• composite socket weld (Du1100);
• primary circuit pipeline and main shutoff valve socket
weld;
• ring of main circuit pump (MCP-310) socket weld.
For example at the acceptance test of Du1100 composite socket
weld was used mock-up with realistic flaws. At the Figures 3 and 4
shown the results of mock-up inspection in search and measurement
modes. At the Figure 4 shown the coherent image of flaw area 1
marked at the Figure 3. In this area were detected two flaws: #1 in
cladding and #2 in the weld root. On the whole all flaws in mock-up
were detected and flaws at the correct positions. Also was detected
unintended flaws, one of this flaws is shown at Figure 3 as red
square at 320 mm coordinate.
The particularity of this weld testing is in use of local track
for scanning device (see Figure 5). The inspection runs in search
and measurement modes for single track position.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
-3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 3,0
Абсолютная погрешность, мм
Количество
измерений
Zверхн ВысотаAbsolute accuracy UT Real
N. o
f me
asu
rem
en
ts
-
4
Figure 3. The result of composite weld mock-up UT inspection
Figure 4. The result of composite weld mock-up UT inspection.
Flaws 1 and 2
Figure 5. Scanning device on the weld
1 2
-
5
At the Figure 6 shown image of real longitudinal crack detected
at the operating
inspection (B-,C-,D-images). White lines shows weld chamfer and
cladding, green lines shoes crack profile. Image was obtained in
measurement mode with transducer emitting from the surface near to
steam generator. Flaw is located at the edge of base perlitic metal
and austenitic cladding. The crack was cut out and its length,
profile and height were confirmed by metallographic
investigation.
Also was measured another crack size in area where amplitude of
indication was below the recording level. At the Figure 7 shown
plots with crack profile measured by AUGUR and by metallography.
Image was obtained in measurement mode with transducer emitting
from the surface near to primary circuit pipeline. Height sizing
accuracy at the obtained samples (50 samples with length 20 mm) was
measured as 4 mm in 95% confidence. It is necessary to precise
accuracy measurement with additional samples and height
measurements for inspection from steam generator side. This
activity is carrying out at this moment.
Figure 6. Coherent image of crack obtained with longitudinal
waves. At the right side template of the
weld shown.
0
5
10
15
20
25
30
1500 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 2050 2100
2150 2200 2250 2300 2350 2400 2450 2500 2550 2600 2650 2700
Координаты несплошности вдоль сварного соединения
Высота несплошности
данные МГИ
Результаты АУЗК АВГУР визмерительном режиме
Figure 7. Flaw profile comparison by UT and metallography.
Weld
Crack
Cladding
UT height
Real height
Position along the weld, mm
Fla
w h
eigh
t, m
m
-
6
Summary and discussion The main work outcome is the creation of
new inspection equipment and techniques,
detection of dangerous flaws in welds. Applied equipment allowed
to describe full and almost exact real weld condition at the NPP in
operation and gave opportunity for future safety forecast.
References
1. Grebennikov V.V., Gurvich A. K., State of art in ultrasonic
testing of austenitic welds,
Defectoskopiya, 1985, № 9, p. 3-12 (in Russian).
2. V. G. Badalyan, E. G. Bazulin, A. Kh. Vopilkin, D. S.
Tikhonov, Application of super-
resolution techniques to expert testing of welded joints in
pipelines of nuclear power
plants. Russian Journal of Nondestructive Testing, 2000 № 1.,p.
45-50.
3. E. G. Bazulin, Increasing the signal-to-noise ratio during
joint use of the spectrum
extrapolation and splitting methods. Russian Journal of
Nondestructive Testing, 2006,
Vol. 42, No. 1, pp. 51-59.
4. V. V. Grebennikov, A. H. Vopilkin, D. V. Grebennikov, V. G.
Badalyan, D. S. Tikhonov.
Ultrasonic testing of austenitic welds: I. Application of
multiparameter methods to
increase efficiency. II. New capabilities with the use of
automated systems. NDT
World., № 1 (19), march 2003, p. 10-15 (in Russian).
5. V.G. Badalyan, E.G. Bazulin, A.K. Vopilkin, D.S. Tikhonov.
Use of Transformed
Waves for Obtaining Images of Vertical Cracks from
Multi-Frequency Acoustic
holograms. Russian Ultrasonics. Vol. 24, no. 2, pp. 75-84.
1994.
6. Krautkrämer J., Krautkrämer H. Werkstoffprüfung mit
Ultraschall. Shringer-Verlag,
1986.
7. Grebennikov V., Badalyan V., Grebennikov D., Vopilkine A.,
Comparative Analysis of
the Ways to Increase Signal to Noise Ratio at NDT Inspection of
Austenitic Welds. 15th
World Conference on NDT, Rome (Italy), 2000.
8. Badalyan V., Vopilkine A. K., The ultrasonic examination and
monitoring of austenite
welds of stainless steel pipelines at Russian Nuclear Power
Plants. 8 Europe Conference
on NDT, Barcelona, 2002.
