ML20086K458
| ML20086K458 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee File:NorthStar Vermont Yankee icon.png |
| Issue date: | 06/20/1995 |
| From: | Kietzman K, Sagstetter M ELECTRIC POWER RESEARCH INSTITUTE |
| To: | |
| Shared Package | |
| ML20086K450 | List: |
| References | |
| NUDOCS 9507200151 | |
| Download: ML20086K458 (47) | |
Text
.
EPRI NDE CENTER Electric Power Research Institute Nondestructive Evafuation Cer\\ter
....... __....-_-_ -._ Leadership in Technology Transfer June 20,1995 Carl Larsen Yankee Atomic Electric Co.
580 Main Street Bolton, MA 01740
Dear Carl,
This is the final report prepared by myself and Kim Kietzman. This report summarizes results to-date on a project to demonstrate ultrasonic examination techniques as a replacement to NUREG 0619 required liquid penetrant (PT) examination of feedwater nozzle inner radius region.
The demonstration was successful in documenting the ultrasonic procedures and equipment considered for use in inspecting the nozzle inner radius region.
Thank you and Bill Fields for your active participation during the demonstration. Your help made the demonstration run more smoothly. Please feel free to call Kim Kietzman l
or myselfif you have further questions or comments.
Sincerely, 1%k S Mark Sagstetter
)
MS:inb vycoviet. doc t
cc:
Dennis Girroir Pat Donnelly Jack Lance l
Frank Ammirato
)
Larry Becker
)
Kim Kietzman utility file j
t l
9507200151 950714 PDR ADOCK 05000271 p
PDR fOgE I Oh #Y7 Telephone. (704) 547-6100 Charlotte. North Carohna 28262 1300 Harfes Boulevard F AX- (704) 547 6168 (P O Dox 217097. Charlotte. North Carchna 28221) 0 L
(
.I l
1 y
c s
o Inside Surface Ultrasonic Technique Capabilities
'As Demonstrated for th'e Inspection of Feedwater Nonles
)
i Final Report Prepared by:
J Mark Sagstetter i
Kim Kietzman This is the final report for the Vermont Yankee / Yankee Atomic (VY/YA) demonstration of ultrasonic techniques for the examination of the feedwater nozzles at Vermont Yankee (VY). This report summarizes the results from the demonstration. Ultrasonic techniques were demonstrated for consideration as a replacement to the NUREG guide 0619 required liquid penetrant (PT) examination of the feedwater nozzle inner radius region.
)
Outside surface examination techniques were initially investigated as a replacement for PT. NUREG 0619 states problems with long metal paths and complex geometry. When it was determined that the VY feedwater nozzle configuration had sufficient access between the thermal sleeve and nozzle to accommodate inside inspection tooling, the emphasis switched to
.l investigating inside surface examination methods. Small gap examination capabilities were recently introduced for other examination configurations such as the CRDM head penetrations, and it was anticipated that examination from the inside surface would deliver an improvement in flaw characterization and sizing because of shorter metal paths and less complicated geometry.
VY/YA contracted with Babcock and Wilcox Nuclear Technologies (BWNT) to develop inspection equipment and procedures to deliver an inspection from the inside surface. The NDE Center assisted VY/YA in developing a demonstration strategy and presenting the proposal to the NRC to obtain relief from PT requirements of NUREG 0619.
Several mock-up configurations and flaw manufacturing processes were discussed.
VY/YA elected to use a combination of EDM notches and weld implanted mechanical fatigue cracks in full scale mock-ups which represented the actual component geometry. Two mock-ups, a practice and a 1
i
demonstration mock-up, were employed for the demonstration. The demonstration was conducted in blind and non-blind phases. The practice mock-up was used for technique development while the demonstration mock-up was used for the individual phases of the demonstration. After technique development on the practice mock-up the blind phase of the demonstration took place. Initial detection and flaw sizing was perfonned in a blind fashion.
Upon completion of the blind phase of the demonstration, grind-outs were placed in the demonstration mock-up to simulate field conditions, and the capability to detect and size flaws in the grind-out areas was perfonned in a non-blind fashion. Local grinding was perfonned on the same flaws used during the blind phase. These grind-out areas and flaws contained in them were representative of the condition experienced at VY.
The NDE Center assisted Yankee Atomic (YA) in administering the demonstration. YA/NDE Center responsibilities included maintaining security of the mock-up during the blind phase, verifying and recording essential variables (see attaclunent 1), and monitoring all phases of the demonstration. Monitoring was especially important during the demonstration because the non-blind portion required an understanding of the equipment and procedures to the level that perfonnance could be assessed.
A fonnal review of the demonstration protocol and BWNT procedures was perfonned prior to the commencement of the demonstration. This protocol was followed during the demonstration process (see attachment 2).
The demonstration provided a means to detennine and document the level of performance of the examination processes and procedures. It also served as a trial nm of the equipment on full scale mock-ups. This enabled an assessment of the equipment and procedures, and allowed for improvements before the actual inspection at VY. The demonstration took place in BWNT's facility in Lynchburg, VA.
l pajf 3 M
Mock-Up Production Two mock-ups were fabricated for the project: a practice mock-up and a final demonstration mock-up. Feedwater nozzles with the same dimensions i
as the VY nozzles were used. A sparger was placed inside the mock-ups when scanning to represent actual clearances realized in the VY configuration i
(see Figure 1).
The practice mock-up contained EDM notches that were placed in the bore region and along the inner radius region. Grind-outs were also included as described below. This mock-up was made available to BWNT for tecimique development.
The final demonstration mock-up contained mechanical fatigue cracks manufactured using a special welding process. The flaw area was prepared by excavating to a proper dimension. A fatigue bar was welded into place and fatigued until it broke free from the weld. The broken piece was then sectioned, shaped to a specific flaw size and fitted into its original position in the excavation. The excavation was then filled with weld metal.
After the flaws were implanted in the mock-up and the cavity was filled, a layer of cladding was reapplied, resulting in subsurface cracks. Although cracks produced in this way are not expected to accurately represent service-induced cracks which would grow by fatigue through the clad, VY/YA i
considered the subsurface cracks conservative because this type of flaw would produce less reflected ultrasonic amplitude than a surface connected crack.
The demonstration mock-up was used for the blind and non-blind phase of the demonstration. After successful demonstration of the techniques during the blind phase, grind-outs similar in size and shape to those found at VY were placed in the mock-up for the non-blind phase. Three grind-outs were placed in the mock-up with the following sizes:.225,.300,.310 inch in depth. The edges were radiused to give a smooth transition to the nozzle surface. The grind-outs were positioned so that the flaw was located at the bottom of the cavity. This was considered to be the most difficult position of the flaws for detection and sizing (see Figure 2).
3 page e of 47
d l
i*
s f
y I
u'
- i f
/
~ ~ ~ ~1
- .4.
g:4 si%i 44,
^ '
kl;th;q r' '.
-t
/
.~
b;
-v E ;7; ' ); y _.Q~
$. 4f j'.,.
.;4?
='
a.Y sucw.aw Y A,ibw.. e <: -i.vwm..w.. ;.< :,. ::
a w a s a. & #3F5 +M W W R & i; ~. M :n.:s
' '"
- Mie 4kN.Qp,..
.+ S1
... m
\\[
W.VY>@th A &j i p,
..-.gl
[].
Q
)!\\l 5l Ik
.I.
- s i
- ~,
.L.
~:
', '/ f,
c4! ' s
, u,:,-
- w. t..
g..,3
\\
Sparger Nozzle inner Radius Nozzle Holder Represents Inside i
Surface of Vessel
-l Figure 1. VY/YA Feedwater Nozzle Mockup
,1 l
i f
I
(
page Sc /11 I
us rcrion o "
~~
1 e Grinc -Ou-- Areas wi-,
aws Nozzle Inner Radius Grind-out Area Flaw in Grind-out Area Figure 2
eoh
-h N
4
Security During the blind phase, the final demonstration mock-up was kept secure at all times. The truth information on the blind mock-up was maintained by YA and the NDE Center. BWNT had access to the examination surface of the mock-up only during the formal demonstration. Security was maintained using locking covers over the ends of the nozzle when the mock-up was not in use and securing the cover of the tank with wire tags and lead security seals while the mock-up was in the tank, but not being inspected.
During the second phase (non-blind phase), the emphasis was on a better TOFD technique, detection, and sizing in the grind-out areas. The second phase was not considered a blind demonstration and the flaw truth information was not kept secure.
4 page 9 of 47
First Phase of Blind Demonstration On November 28 through December 8,1994, the blind demonstration took place. This was the first phase of the two part demonstration. The purpose of this phase was to demonstrate the initial detection / characterization capabilities of BWNTs ultrasonic procedures on the inner radius region before grind-outs were placed in the mock-up. The second phase would i
consist of scanning the mock-up again for detection and sizing of the flaws in the grind-out areas.
4 i
During this phase of the demonstration it was BWNTs intention to i
demonstrate both fonvard and backward scatter tip diffraction techniques on the blind mock-up. BWNTs procedure employed high angle (80-degree),
dual-element, longitudinal-wave transducers for detection of the flaws and two 52-degree, dual-element longitudinal-wave transducers, positioned in a fonvard scatter arrangement for flaw sizing. BWNT intended to operate the 52 degree transducers in either a backward or forward scatter mode.
100% detection was achieved before the grind-outs were placed in the mock-up. Data from both fonvard and backward scatter tip diffraction techniques v'as used for the determination of flaw size. This resulted in an RMS error of 0.11 inch (see Figure 3). BWNT did not produce sizing results on the techniques independently. Flaw sizing performance improved when the backward scatter data was analyzed independent of the fonvard scatter data, indicating a problem with the fonvard scatter procedure.
Backward scatter data was used to identify the flaw location and confirmation was attempted with the fonvard scatter technique. The process of using a combination of backward and fonvard scatter tecimiques in this fashion is desirable, however BWNTs fonvard scatter procedure did not deliver the accuracy anticipated.
5 page 8 of 47
a
=
t s
E e
s.
n v
it iL F
h l
t t
a u
s e
r e
d T
B I
8 3
9 9
5 0
=
1
=
=
1
=
d 1
e f
r e
o v
o
=
v r
e r
E D
C e
r S
s r
M b
d d
r t
t o
O S
S C
R 0
000 1
ooo s.o S
/
TLUS E EC 3
RA o
F o
e o
r G R P s.
u N UU o
ig F
IZS
)
K n
ISD C i(
H AO H
L T
TP C M R
U E ED T
D HN W TL 0
I 0
B 0
AM 1
4
/
o LO F R
/
/F N
W
/
F 0
e 00 2
0
/
,//
0000 0
o o
o o
o o
o o
o o
o o
o.
s.
4 o.
1 o
o g
o
= 8mS6E Ddt
%49
I When using the fonvard scatter technique on cladded materials it is expected that the lateral wave travels at the clad to base metal interface. The lateral wave should be relatively constant, only being interrupted by a flaw which either breaks the surface of the clad and extends into the base material or where the tip of the flaw is at the clad to base metal interface and the flaw extends into the base material. The intermption in the lateral wave is then used as an indicator to locate the bottom tip of the flaw. BWNT's fonvard scatter data showed the lateral wave response interrupted in several areas when using two 52 degree dual element transducer operating in TOFD mode.
Therefore the lateral wave could not be used as an indicator as to the flaw tip.
because of the high frequency ofinterruptions in the lateral wave and other signal responses that could be confused with the flaw tip response.
The unsatisfactory perfonnance could be attributed to either the transducers, the process used to fill the cavity or cladding application. The transducers are dual element transducers operating in a TOFD mode. Both elements of the sending transducer were pulsed simultaneously and the receiving transducer receives with both elements simultaneously. TOFD is usually done with a single element transducer and using dual element transducers in this fashion might have affect perfonnance.
A discussion with the flaw manufacturer revealed that the clad metal was first applied in a circular motion and then in a back and fourth motion across the flaw. When the nozzle was originally cladded, the cladding was applied in the nonnal fashion following the circumference of the nozzle. This unusual clad condition may also have affected the perfonnance of the fonvard scatter l
technique.
Since the flaw tip responses were clearly present in the backward scatter data i
and there was no confusing flaw tip responses, the backward-scatter technique was not affected. Therefore the detection of the flaws was not affected.
The final phase of the demonstration, placing grind-outs in the inner radius region for demonstration of detection and sizing capabilities, was postponed, at the request of VY/YA. This allowed extra time for BWNT to obtain new transducers in an effort to improve the time of flight fonvard scatter technique and for flaw modifications.
f*R /0 Oh Y 6
i l
VY/YA initially chose to use subsurface flaws in the mock-up to be conservative in their evaluation of BWNT's technique. However since the procedure did not deliver acceptable results for both Forward and Backward l
scatter techniques, VY/YA elected to modify the flaws by making the flaws surface connected. Electro Machined Discharged (EDM) notches were l
placed above the flaws to represent the surface-connected ligament of the flaw.
Locating and positioning the flaws for the EDM process was accomplished using an 80-degree, dual-element, longitudinal-wave transducer. When grind-outs were later placed in the mock-up, it was observed that these notches lined up very well with the original flaws.
l 7
i gage // or 47
l l
Detection Results First Phase High angle (80-degree) dual element longitudinal-wave transducers were used for flaw detection. The transducers were arranged in the standard pulse-echo mode. Scans were made in segments around the circumference. The major scan axis was circumferential, incrementing in the axial direction.
Transducers were oriented to detect axially-oriented flaws only. Information obtained from the detection scan was also used for flaw length determination and a qualitative estimate of the flaw through-wall depth.
Detection performance on the blind demonstration mock-up was 100%.
i Flaw Position Results First Phase Preliminary analysis of the results showed an RMS error of.25 inch in the determination of flaw axial position. Additional regression analysis results are shown in Figure 4. Since this enor was consistent on every flaw, it was 6
discussed with BWNT. BWNT was then able to identify a systematic position error. After the error was identified and corrected, the axial positioning error was detennined to be.19 inch (see Figure 5)
The RMS error for circumferential flaw positioning was 3.3 degrees. Results are shown in Figure 6.
8 pgef.Lof47
E FWN / AXIAL LOCATION OF FLAW BLIND MOCK-UP 4.000 Std Dev =.69
/
//
Std Error =.16 3.000 f
/
"E
=
z 52 Corr Coef =.98 t-m 2
s RMS =.25
<C 2.o00
/
e
/
[
- Observed =5 m
/
6
/
/ /
Truth vs. Est.
i Best Fit N
+
l p
b ideal Line m
./
i i
/
b
/
l C>
0.000 i
i g
0.000.
1.000 2.000 3.000 4.000
~k.
Q TRUTH AXlAL POSITION (in)
- Before B&W made changes to the axiallocation measurements Figure 4
_ _ _ _ _ - - _ _ _ _, - _ _.. _ _. _ -. ~...
i l'
FWN / AXlAL LOCATION OF FLAW BLIND MOCK-UP 1
l 4.o00 1
Std Dev =.69 i
i 3.o00 Std Error =.16 3
1 4
t l
z i
9 Corr Coef =.98 t-u>
i On.
3 RMS =.19
< 2.000 l
R
/
8
/
- Observed =5
/
y
/
1.000
-[
Truth vs. Est.
.i Best Fit i
/
ideal Line (R
/-
wk o.ooo -
6 h
0.000 1.000 2.000 3.000 4.000 '
A TRUTH AXIAL POSITION (in)
N
- After B&W made changes to the axiallocation measurements Figure 5
-... ~..
FWN / CIRCUMFERENTIAL LOCATION RESULT BLIND MOCK-UP 400.0
'/.
8 Std Dev =80.51 Std Error =3.35 0
/
E
/
i o
/
E
/
Corr Coef = 1.00 z9w
@ 200.0.
O RMS =3.29 degrees 8
cc 0
- Observed =5 g
2 100.0 Truth vs. Est.
Best Fit 4
ideal Line s
0.0
[
1 0.0 100.0 200.0 300.0 400.0 t
TRUTH POSITION DEGREES Figure 6
ts E
e s
n v
it iL F
h la t
t u
s e
r e
T B
d I
8 9
5 6
1 6
1
=
=
=
5
=
d r
e 2
e f
v v
o o
=
r e
r C
e r
D E
S s
rr b
d d
o M
O t
tS C
R 0
00 3
00 5
2 ST L
o U
o o.
S
./
E 2
7 R
e
/
G ru P
N g
U
/
iF IZ
-K
)
/
i I
n S C
(
0 HO 0 H 5 T TM
/
U G
1 R
ND
/
T EN LI
/
L WB A
L F
0 0
/
0 1
N W
/.
F
/,
005
/
o
/
000 0
o 0
0 o
o 0
0 0
0 o
o 0
3 2
2 1'
~
0 0
5 0
S o.
1 0
_E8CS<$
%s g
so 3
Flaw Depth & Leneth Sizine First Phase j
1 Sizing scans commenced after the completion of the detection scans and j
detection analysis. Sizing was performed using (52-degree), longitudinal-wave, dual-element, transducers operating in both a forward and backward scatter modes.
BWNT calculated the through-wall dimension of the flaw using two methods:
- 1) locating the diffracted signal from the bottom of the crack and determining 7
its position relative to the clad surface, and 2) locating the bottom tip relative to the clad-to-base metal interface. Calculation relative to the clad surface.
l resulted in an RMS error of.11 inch, as shown in Figure 3.
Figure 7 shows flaw length sizing results. The resulting RMS error was.25 Inch.
i
?
i i
i 9
fy l'l 0l Yl
Second Phase In February of 1995, the demonstration reconvened for the purpose of demonstrating improved TOFD techniques with new transducers. The flaw truth information was not revealed to BWNT at this time, however because the RMS errors from the first phase of the demonstration were made available to BWNT, this phase of the demonstration was no longer considered blind.
As the demonstration neared completion of the second phase, flaw truths became more evident to BWNT.
The new transducers were tried before the grind-outs were placed in the mock-up, three separate techniques were used for sizing the flaws. The first technique consisted of a Sigma 52 degree dual element transducers operating in TOFD mode. The second technique consisted of KBA 52 degree dual element transducers operating in TOFD mode, while the third technique consisted of KBA 52 degree single element transducers operating in TOFD mode. The comparison of these techniques revealed that the KBA single element transducer gave the lowest RMS error.
Grind-outs were then placed in the demonstration mock-up along the inner radius region. The mock-up was then scanned again for detection of the flaws with the 80-degree dual element longitudinal-wave transducer. This technique was unable to clearly detect all the flaws in the grind-out areas.
A new technique was attempted consisting of a 52-degree transmitting transducer and an 80-degree receiving transducer in a backward scatter mode.
This will be referred to as the Tandem techniaue (see Figure 8). The Tandem l
technique showed improved detection capabilities in the grind-out areas.
l paqi /9 00 41
U S"rQ"l O n O"
Q n C e rn ec inic ue i
Sound induced to excite flow
/-
8 Fla w f'N
/
s, N
k
- eld 80 degree 52 degree Receiving Sending
/
Transgucer Tra ns<jlucer j
/'
/
x
/
/
NJ
/
/
s 9
/
/
s Diffracted sound form flow lip T
j Figure 8
O b
-t s
~
Detection in Grind-outs Second Phase of Demo During the second phase of the demonstration the 80-degree longitudinal wave transducers were used for detection. This technique demonstrated 100% detection before grind-outs, but was unsuccessful in the ground-out areas.
BWNT then introduced the Tandem tecimique for detection in grind-outs.
This tecimique gave better detection results than the 80-degree longitudinal wave transducer by itself. The Tandem technique gave 100% detection of the flaws in the grind-out areas and 100% detection of the remaining flaws not in grind-out areas.
Positionine Results The flaw positioning results are taken from the first phase of the demonstration and are considered to have an axial position RMS error of 0.19 inch.
The RMS error for circumferential flaw positioning was 3.29 degrees. (see Figures 5 & 6)
Depth Sizine Results before Grind-outs Several techniques were tried and evaluated separately. This was done in an efTon to improve on the TOFD results (see Figures 9 - 11).
Sigma Dual Element sizing results RMS error.13 inch i
KBA Dual Element sizing results RMS error.11 inch KBA Single Element sizing results RMS enor.09 inch Length Sizine Results before Grind-outs The scan increment did not change so length sizing results are expected to be the same as demonstrated in the first phase of the demonstration see Figure 7.
\\
p aj e.2 0 0 A A 11
4 i
TOFD SIGMA DUAL ELEMENT SIZING RESULTS FROM THE CLAD SURFACE 1.ooo Std Error =,13 Std Dev =.13
/
z Corr Coef =.67 y
=
/,
y o.4oo
/
/
- Observed =3 Truth vs. Est.
o.2co g
P Best Fit t
ideal Line b
I D
- o. coo h
4 o.ooo c.2oo o.4oo o.soo o.soo 1.000 M
TURTH (in)
- Before Grind-outs Figure 9 t
.,,, - - - -... -,...., - -,, _... ~.,,. ~
--.--.n.
, ~, - -.. -., -.., - - -, -.
t E
s e
s n
v it i
F L
h l
S t
t E
u s
a H
r e
e T
B d
2 9
C J.
i N
3 0
9 I
=
0 1
=
=
2
=
d 1
e f
r e
o v
o
=
v r
e r
E D
C e
r S
s r
M b
d d
r t
t o
O S
S C
R 0
0 0
1 0
j l
0 8
0 T
E N
C 0
E A
1 0
M F
0 e
r 6
u ESR
/
i 0
g LTU i
F ELS
)
U n
L D
S
(
A EA H
U RL T
D C
R G
U E
A N T
BI H K ZT 0
IS 0
D M
4 0
F O
O R
T F
0 02 0
///
/
s 0
t 0
u 0
o O
d o
O 0
o 0
0 n
c o
0 o
0 0
i o.
s.
6 4
2 0
G r
o 0
o 0
o 1
e r
i w _2 o
fe B
- NbbOwA4, 3
r 1
TOFD KBA SINGLE ELEMENT
[
SIZING RESULTS
.j FROM THE CLAD SURFACE 1.ooo i
Std Error =.08 o.800
/
Std Dev =.16 i
/
/
/
- o.6oo
-[
Corr Coef =.94 4
o i
1 E
RMS =.09
+
INCHES 2
-/
o'4oo
~
/
- Observed =3 Truth vs. Est.
i o.2oo r
p Best Fit ideal Line
/
o.000 o
1 D
o.ooo o.2oo o.4oo o.soo o.soo 1.ooo t
Q TURTH (in)
- Before Grind-outs Figure 11
Denth Results after Grind-outs e
After the grind-outs were placed in the mock-ups, the Tandem technique was introduced due to the problem of detecting the flaws in the grind-outs with the 80-degree longitudinal-wave transducer technique. The Tandem technique was then compared with the other techniques for sizing in grind-out areas (see Figures 12 - 15).
Sigma Dual Element (Backward Scatter)
RMS error.19 inch i
KBA Single Element (TOFD) and t
Tandem Technique combined RMS error.08 inch Tandem Technique only (TOFD)
RMS error.08 inch KBA Single Element only (TOFD)
RMS error.12 inch Lenuth Sizine Results after Grind-outs j
The scan increment did not change so length sizing results are expected to be the same as demonstrated in the first phase of the demonstration see Figure 7.
P 12 page.2 Y Ob YY
4-BACKWARD -SCATTER SIGMA DUAL ELEMENT SIZING RESULTS FROM THE CLAD SURFACE 1.ooo i
Std Error =.09 0.800
/
Std Dev
.19
,/
Corr Coef =.93
- 0.600 i-
/
a 3
/
RMS =.191NCHES
'/
6 2 o.400
- Observed =4
/
/
g,gg Truth vs. Est.
3
[
Best Fit h
/
ideal Line
/
o.ooo g
h O.000 0.200 0.400 0.600 0.800 1.000 h
TURTH (in)
- After Grind-outs
. Figure 12
.,. - ~. -. - -., -. -. - -... -. _ ~..
e TOFD & TANDUM TECHNIQUES COMBINED -
SIZING RESULTS FORM THE CLAD SURFACE 7'
1.000 Std Error =.08 o.Soo
/
Std Dev =.21 p
/
Corr Coef =.96
- o.soo e
/
/
RMS =.08 INCHES o.4oo
/
- Observed =3 Truth vs. Est.
o.200 Best Fit-ideal Line 6
Q
/, /
o.ooo g
-t o.ooo o.2oo o.4oo o.soo o.soo 1.000 TURTH (in)
- Af ter Grind-outs Figure 13
. - ~
=-
TANDUM TECHNIQUE SIZING RESULTS FROM THE CLAD SURFACE
' %0
/
Std Error =.10 f
i o.soo Std Dev =.15
/
/
Corr Coef =.86 o.soo
/
.E
/
O
/
y a
RMS =.08 INCHES
+
o.4oo
/
- Observed =4
,/
I
/ /
Truth vs. Est.
o.200 Best Fit l
9t Ideal Line g
p d
'/
o.ooo g
h 0.000 0.200 0.400 o.600 0.800 1.000 t
TURTH (in)
-q
- Af ter Grind-outs Figure 14
TOFD KBA SINGLE ELEMENT SIZING RESULTS FROM THE CLAD SURFACE t.o00 Std Error =.03 o.soo Std Dev =.27 Corr Coef = 1.00
- 0.6o0 i
/
w RMS =.12 INCHES o.4co p
/
- Observed =3
/
/+
+
Truth vs. Est.
o.20o r
Best Fit w
Ideal Line
- G
,/
- o. coo 4
o.000 o.2co 0.4oo o.soo o.800 1.000 4
TURTH (in)
- After Grind-outs Figure 15
O Data Review by EPRI NDE Center. BWNT. VY/YA Personnel After reviewing the results, BWNT proposed to use the 80 degree dual element technique for detection and the 52 degree technique for sizing in the l
bore region and the Tandem technique for detection and sizing in the inner radius region.
YA/NDE Center review of the data revealed that the sizing methodology used with the tandem technique was to pick the position showing the highest amplitude response from the flaw tip and to calculate flaw depth form this position. This is an acceptable practice. However, another method was suggested and proved to be more accurate. This method involved observing the echo dynamics from the flaw and detennining tip position based on both amplitude and echo dynamics. The Tandem Tecimique (Envelope sizing method) demonstrated an RMS error of.030 inch (see Figure 16).
Bore Region (80-degree detection /52-degree sizine)
BWNT proposed using the 80-degree dual element longitudinal-wave l
transducers for detection and a 52-degree dual element fongitudinal-wave transducers operating in a TOFD technique for sizing the flaws in the bore i
region. This was based on the results of the practice mock-up (the demonstration mock-up contained no flaws in the bore region) However the data from the practice mock-up proved the tecimiques to be viable for this region.
l t
I l
13 ya.9 6.f 47
q L
e ts E
e s
n v
it i
F L
h S
t t
la E
u s
e r
e H
T B
id 2
9 C
6 9
N 5
0
=
=
I
=
1 3
=
d 0
e f
r e
o v
o
=
vr r
e E
D C
e r
S s
r M
b d
d r
t t
o O
S S
C R
0 0
o.
~
1 0
+
0 8
0 EC 6
1 E
A o
U F
o e
s.
u r
QSR j
ig o
)
HU CSD in
(
EEA H
TRL T
C R
MG U
U NE
.T I
H DZ T 0
I N S 0
A M
4 0
T O
)
d R
o F
h te m
ep
/
+
loe v
0 n
0 E
2
/ /
is
(
0 gn u
g n
e iz is dn a
/
o s
c t
o.
u o
o 0
o 0
o o
d 0
o 0
c o
n o.
s.
4 2
o.
ir 1
o o
o o
G r
= 8 :a 3 6 2 e
t oh44 A
f
%4%
Inner-Radius Technione (Tandem Technione)
Detection The Tandem technique was proposed for detection and sizing the flaws in the inner-radius region of the mock-up. Although the technique demonstrated i
100% detection and respectable sizing capabilities in and out of the grind-out areas, there are a couple of conditions not accounted for that exist in the plant. The first condition could affect detection and sizing when two grind-outs are closely spaced in the inner radius. However, this condition is known to exist in only four possible locations (two each in two nozzles) at VY. As a comparison, the scan area reduction in coverage due to the spacer weld pad interferences was a maximum of 2.0 percent on each nozzle. The maximum reduction in coverage due to closely spaced grind-outs on one nozzle would add an additional.9 percent and on the other nozzle an additional.6 percent (see Figure 17).
Sizine During the "non-blind" demonstration the bottom tip of the flaw was identified using the Tandem technique. There is no tip signal from the top of the flaw because it is surface connected. During the demonstration the position of the transducer and the depth of the grind-out were known. In actual practice this would not be the case. This exam uses backward-scatter principals and flaw depth (thru-wall) is detennined based on transducer position. As seen in figure 6 circumferential positioning error was as great as 3.29 degrees. This would affect the transducer position relative to the flaw, which in turn would affect thru-wall dimensioning of a flaw (See Figure 18).
However, further evaluation of demonstration could provide more infonnation on thru-wall dimensioning. The 3.29 degree RMS error is relative to the nozzle coordinates. During the demonstration the flaw target motion response could be related to the grind-outs which display their own distinctive responses. The relationship between the flaw response and the grind-out response could be compared and the actual sizing error due to error in positioning with in the grind-out could be detennined in the demonstration data if necessary.
fYdI 14
S 't " C " l O
^
O
J, e e c,t.io l
G~^c-ot-A eas.
Inner Radius Transducer Set-up e
lM Side by Side Grind-outs,/
Flow 3d~g
-Ic u re 7
o ty
6 t s rc~ ion of 3 0ssia e E ~or 'n Jeom Re atec to Er or in P o si-io n l
Sca e
=
o Nozzle
/
Inner Radius j/
l lilustrates Position Error l
in degrees (RMS 3.29 degrees) f' l
\\
/
's
/
r.310" o
N N
Possible Depth error j
d Illustrates deepest grind-out Related to error in p o sit,io n.
worst case senario I
j rig u re
'8 4
Conclusion Two mockups a practice and a demonstration mock-up were fabricated for the demonstration. BWNT used the practice mock-up for procedure development and the demonstration mock-up for a blind demonstration of equipment and procedures.
The demonstration was broken up into two phases. The first phase was to demonstrate UT detection and sizing on the blind demonstration mock-up before grind-outs while the second phase demonstrated technique capabilities after the grind-outs.
After the first phase of the demonstration, the demonstration was no longer considered blind. A significant amount of procedure development took place during the demonstration requiring modification of the original demonstration plan. BWNT's original procedure was modified significantly. However, the demonstration was successful in documenting capabilities of the final BWNT procedure. This included a detection technique which delivered 100%
detection and a sizing technique which delivered.11 inch RMS error for the bore region.
Originally, BWNT planned to inspect grind-outs by placing transducers directly in the grind-out. Due to the difficulties with this concept, a tandem technique was introduced for scanning the inner radius region. This technique as described on page 10, was used for detection of flaws along the inner radius region and delivered 100% detection of flaws not in the grind-out areas and 100% detection of the remaining flaws in the grind-out areas. However, detection could be afTected if grind-outs are closely spaced although the area affected by this scenario is minimal.
Sizing with the tandem technique initially delivered an RMS error of.080 inch. After review of the data another sizing methodology was suggested which delivered an RMS error of.030 inch. Even though the RMS error was improved upon, in practice the error could be greater due to unknown dimensional inaccuracies related to the grind-outs in the plant. (as described on page 14).
page 34 6b N9
VERMONT YANKEE l BABCOCK AND WILCOX NUCLEAR TECHNOLOGIES RPV Feedwater Nozzle Demonstration (Essential Variable, detection and sizing, Axial Flaws)
Instrument or System ACCUSONEX Data Acquisition and Imaging system see section 5.0 for more information Digitization Frequency 50 MHz j
Procedure Remote Underwater Ultrasonic Examination for Detection and Sizing of Axial Flaws in BWR i
Feedwater Nozzles.-ISI Procedure NO.116 REV.0 Search Units for Detection Search Unit identification ( 80 degree Sigma 80 dualcrystal i
technique )
Search unitTandem technique Sigma 80 dualcrystal Sigma 52 dualcrystal i
Center Frequency 2.25 MHz Bandwidth or Wave form duration 80 = 50 %,52 = 80%
Mode of Propagation Longitudinal Nominal inspection Angle (80* technique) 80 Nominal Inspection Angle (Tandem 52 sending 80 receiving technique)
Probe Center Separation (Tandem technique) 12 Number of Elements (Tandem technique) 2 Sending,2 receiving Number of Bements (80* technique) 2 Search Units for Sizing Search Unitidentification (52 technique)
Sigma 52' dual crystal i
Search Unit identification (Tandem technique)
Sigma 80 dualcrystal Sigma 52' dualcrystal l
Center Frequency 2.25 MHz / 5.0 MHz j
Bandwidth or Wave form duration 80 = 50 %. 52* = 80%
Mode of Propagation Longitudinal Nominal Inspection Angle (Tandem 52 sending 80 receiving technique)
Nominal Inspection Angle (52 technique) 52 Probe Center Separation 12 Number of Elements (Tandem technique) 2 Sending. 2 receiving Numberof Elements (52 technique) 2 j
pa3c35o H 7 1
VERMONT YANKEE l BABCOCK AND WILCOX NUCLEAR TECHNOLOGIES RPV Feedwater Nozzle Demonstration (Essential Variable, detection and sizing, Axial Flaws) t Element Size Detection and Sizing Size of Elements
.40' x.20' dual l
Shape of Elements Rectangle 1
Search Unit Cable Type See Section 5.2 for information on all cabling Maximum Length Maximum # of Connectors Detection and Sizing Technique Scan pattern and beam direction See Scan Plan l
Maximum scan speed 2" per second Pulse Repetition Rate See 8.6 of the Procedure l
Extentof Scanning See Scan Plan Methods of Calibration for detection and sizing See Procedure inspection and Calibration data to be recorded i
Method of data recording First recorded onto the computers internal hard drive then copied to an optical disk Recording equipment Color printer, optical disk Method and Criteria for the Discrimination See section 9.0in the procedure o' Indications l
l Surface preparation requirements l none reauired for demonstration l
i I
i page 39 oM1 1
\\
EPRI NDE CENTER Electric Power Research Instdute Nondestructrve Evaluatson Center Leadershipin Technology Transfer PROTOCOL FOR DEMONSTRATING THE PERFORMANCE OF FEEDWATER NOZZLE EXAMINATION PROCEDURE YANKEE ATOMIC ELECTRIC COMPANY / VEBf.IONT YANKEE VY / EPRI NDE Center 10/11/94
/O[l7[9T Prepared by:
f Date:
/
/B//7!Y b"
Date:
Reviewed by:
r Approved by 1/>A Date: /d-I-Yf by Mark Sagstetter 10/17/94 Te6ephone:(704) 547 6100 g[f[
1300 Harris Doulevard Charlotte, North Carolina 2E262 a
(P.O. Box 217097, Charlotte. North Carolma 28221)
F AX. (704) 547-6168
i o
TABLE OF CONTENTS 1.0 GENERAL 1.1 Mock-up Description 1.2 Flaw Description 2.0 EXAMINATION PROCEDURE 2.1 Essential Variables 2.2 Procedure Requirements 3.0 ESSENTIAL VARIABLES 3.1 Ultrasonic Essential Variables 4.0 DEMONSTRATION PLAN 5.0 PERSONNEL REQUIREMENTS t
6.0 EXAMINATION GROUP COMPOSITION 6.1 PersonnelFunctions 7.0 DEMONSTRATIONPROCESS Blind 7.1 Detection Phase 7.2 Analysis Phase 8.0 DEMONSTRATION PROCESS Non-Blind 8.1 Detection Phase 8.2 Analysis Phase 9.0 RESULTS REPORTING 9.1 '
Flaw Detection 9.2 Flaw Location Accuracy I
d.
l 93 Flaw Sizing l
10.0 - RE-TESTING 11.0 DOCUMENTATION 12.0 SECURITY 1.0 GENERAL This document will serve as a guideline for the demonstration of examination processes, l
procedures and equipment for the in-service ia=amion Vermont Yankee (VY) feedwater nozzles. This document pertains to both blind, and non-blind demonstrations. For non-blind demonstrations, the truth information will be available to the vendor. During the i
blind demonstration the tme flaw size and location will be maintained by the performance l
demonstration administrator (PDA). The performance demonstration will be conducted in a manner that addresses all pertinent exam parameters of the actual ISI. The performance l
demonstration is intended to assure the reliable detection, location and sizing of flaws j
during feedwater nozzle examination.
VY, with assistance from the EPRI NDE Center, will serve as the performance demonstration administrator (PDA). Duties in this administrative capacity will include responsibility for both blind and non-blind demonstrations. These responsibilities include:
performance demonstration protocol development, ISI vendor examination procedure l
review, demonstration monitoring, demonstration results reporting, specimen security, and maintenance of the documentation generated during the performance demonstration.
i In advance of the demonstration, the activities that will be performed to demonstrate the j
designated examination system and procedures will be described in a document developed by the vendor. The document will be referred to as a " Demonstration Plan". The l
demonstration plan will address detailed compliance with this document. A copy of the l
examination procedure and the demonstration plan will be provided to the PDA prior to i
the commencement of the demonstration.
l Acceptance criteria will be as specified in VYs' formal NDE Bid Specification. VY will bc responsible to determine whether the ISI procedure is adequate. Flaw detection, location and sizing results will be determined upon completion of data collection and analysis. To insure the credibility of the demonstration process, the Vendor's procedure must contain definitive steps for identifying flaw signals and sizing the flaw dimensions which can be monitored and documented during the demonstration. These steps will be discussed with the PDA in detail during both blind and the non-blind demonstrations. The NDE Center l
will assist in evaluating the results of the demonstration.
1.1 Mock-up Descriptions The blind and non-blind demonstration mockup will be a full-scale representation page3 9 of h'7
t i
i of the VY design. The mock-ups will be made from production materials and will l
be mounted in a representative orientation.
l r
P l
l 1.2 Flaw Descriptions i
Mar =#=cenrod flaws will be employed in the blind demonstration mockup. These flaws will be manufactured fatigue cracks introduced on the inside surface of the nozzle. These flaws will be orientated in the axial direction.
j Following the blind demonstration grind-outs will be employed in the j
demonstration mock-up, in the areas of the manufactured fatigue cracks.
l i
Additionally, EDM notches will be employed in a seperate non-blind demonstration j
mockup.
l 2.0 EXAMINATION PROCEDURE i
The Vendor's procedure will include the following.
i i
2.1 Essential Variables Clearly identified essential variables. Unless otherwise stated in this document, l
the examination procedure shall identify parameters for the essential variables i
defined in Section 3. Essential variables will be specified by a single value or a range of values in the examination procedure and detailed in the demonstration plan.
l l
2.2 Procedure Requirements l
Certain information is required to address security and insure the validity of the j
performance demonstration activity. The information specified in 2.2.1 shall be made available to the PDA prior to the commencement ofthe demonstration. Any j
information considered to be sensitive or confidential will be treated as such, and l
will be returned to Vendor upon completion of the demonstration.
2.2.1 F-amination system description.
1 2.2.1.1 System operation maan=1 j
2.2.1.2 Description of all permanent and temporary system storage devices.
pay 40 o/ 47
.e-2.2.1.3 System software revision number.
2.2.2 The Vendor's standard detection and analysis examination result reports will be employed. Additionally, data will be traiwil,ed by the Vendor to a demonstration-specific report form.
3.0 ESSENTIAL VARIABLES 3.1 Ultrasonic Essential Variables 3.1.1 The examination procedure shall contain a statement of scope that specifically defines the limits of procedure applicability (e.g. materials, thickness, diameter, product form).
3.1.2 The examination procedure shall specify a single value or a range of values for all of the identified essential variables.
3.1.3 The examination procedure shall specify the following essentialvariables:
3.1.3.1 Identification of the examination system and ultrasonic instrumentation. The examination system description shall include the software revision number. The ultrasonic instrumentation description shall include the===A*arer, model and series of pulser, receiver, and amplifier.
3.1.3.2 search units, including:
(a) center frequency and bandwidth or waveform duration; (b) mode ofpropagation and nominal inspection angles; (c) number, size, shape and configuration of active elements and wedges or shoes; 3.1.3.3 search unit cable, including; (a) type; pagediof49
L e
4 (b) maximum length; (c) maximum numiser ofconnectors; i
i i
i 3.1.3.4 detection and sizing techniques, including:
(a) scan pattern and beam directions; (b) maximum scan speed; j
(c) minimum and maximum pulse repetition j
rate; (d) minimum sampling rate (automatic
)
recording systems);
l (e) extent of scanning and action to be taken l
for access restrictions-
{
3.1.3.5 methods ofcalibration for detection and sizing l
(e.g. actions required to 'msure that the l
sensitivity and accuracy of the signal amplitude and time =*=+s of the examination
)
system, whether displayed, recorded, or automatically processes, are repeated from i
one examination to the next examination.
l 3.1.3.6 inspection and calibration data to be recorded; (a) method ofdata recording; J
(b) recording equipment (e.g., strip chart, analog tape, digitizing) when used; 3.1.3.7 method and criteria for the discrimination of indications (e.g., geometric versus flaw indication and for length and depth sizing of flaws);
3.1.3.8 surface condition requirements.
4.0 DEMONSTRATION PLAN g4 Yd 0b Y7 i
O The Vendor activities that will be performed to demonstrate the capabilities of their designated examination system and procedures will be described in the demonstrat on plan.
i The plan will address in detail, compliance with this document. A copy of the demonstration plan and the examination procedure will be provided to the PDA for review prior to the commencement of the demonstration.
5.0 PERSONNEL REQUIREMENTS The Vendor will be required to provide documentation of candidate certification and training compliant with applicable Codes and the examination procedure. This information will be included in the Demonstration Plan document as specified in Section 4.0.
6.0 EXAMINATION GROUP COMPOSITION The Vendor's procedure shall identify the responsibilities and qualification requirements for personnel carrying out the following functions 6.1 Personnel Functions 6.1.1 Scan plan development, examination system setup, calibration and data acquisition.
6.1.2 Reviewing acquired data and screening data for detections 6.1.3 Flaw charactedzation and sizing.
7.0 DEMONSTRATION PROCESS Blind Demonstration The Vendor will be required to demonstrate examination effectiveness on appropriate test specimens. The test specimens will consist of two types of mock-ups a non-blind and blind. The blind mock-up will contain intentional defects. During the blind demonstration permanent specimen identification and flaw locations obscured at all times. After the blind demonstration with this mock-up, this mock-up will be machined to have the appropriate sized grind outs placed in the mock-up. The grind-out locations will be made available to the vendor for demonstration of the vendors capabilities to examine in the grind out areas.
Access to the specimens will be limited to the inside examination surfaces only.
The demonstration will consist of two phases; detection and analysis. No time limit will be imposed. All examinations must be successfully completed prior to disclosure of g e#3O M
4:
performance results.
7.1 Detection Phase The detection phase will be performed in strict accordance with the Vendor's formal procedure as specified in section 2.0. The region to be examined will be identified by the PDA.
7.1.1 Calibration, acquisition, and data review steps will be performed in accordance with the vendor's procedure.
7.1.2 The monitor may at any time during the detection phase, request an explanation or demonstration of a procedural step.
7.1.3 Upon completion of the detection phase, the detection results, reports will be completed.
7.1.4 A copy ofdetection results and a copy of all acquired data generated will be transferred to the PDA.
7.2 Analysis and Flaw Characterization Phase The analysis and flaw characterization phase will be performed in strict accordance with tle Vendor's formal procedure as specified in section 1.0 7.2.1 The analysis phase will be performed by the appropriate personnel as identified in section 6.0.
l 7.2.2 Detection data to be analyzed for sizing will be identified by the PDA. Other pertinent information may be requested ifit is i
specifically identi6ed in the Vendor's procedure and the i
demonstration plan.
l 7.2.3 The vendor will fully explain the steps needed during the analyse phase. The monitor may at any time during the analysis phase, request further explanation or demonstration of a procedural step.
7.2.4 The monitor may at any time during the analysis phase, request an explanation or demonstration of a procedural step.
7.2.5 Upon completion of the analysis phase, formal analysis results i
i page 44 0f 41
+
- /
report willbe completed.
' i 7.2.6 The formal analysis report and all acquired data l
will be transferred to the PDA.
7.2.7 Re-looks or re-examinations may be performed as specified in the formal procedure.
.i 8.0 Demonstration Process Non-Blind Demonstration l
s During the non-blind portion of the demonstration procedural steps m detection and analysis will be demonstrated. The non-blind mock-up will contain EDM notche type intentional defects and grind out areas representative of the respective nozzles. With the permanent specimen identification and flaw locations available to the vendor.
i The blind demonstration mock-up will becone past of the non-blind demonstration process after it has the apropriate grind-outs machined into it, as described in l
section 1.2.
8.1 DetectionPhase f
8.1.1 Calibration, acquisition, and data review steps will be performed in accordance with the vendor's procedure.
8.1.2 The vendor will fully explain the steps needed during the analyse phase. The monitor may at any time during the detection phase, request an explanation or demonstration of a procedural j
step.
l 8.1.3 Upon completion of the detection phase, the detection results, reports will be completed.
8.1.4 a copy of detection results and a copy of all acquired data generated will be transferred to the PDA.
8.2 Analysis and flaw Characterization Phase i
The analysis and flaw characterization phase will be performed in strict accordance with the Vendor's formal procedure as specified in section 1.0.
e 450M
.,-,a m
<--mm
ja.
8.2.1 ' The analysis phase will be performed by the appropriate personnel as identified in section 6.0.
8.2.2 Detection data to be analyzed for sizing will be Identi6ed by the PDA. Other pertinent information may be requested ifit is ph=Hy dentified in the Vendor's procedure and the j
demonstration plan l
8.2.3 The vendor will fully explain the steps needed during the analyse phase. 'Ihe monitor may at any time during the analysis phase, request further explanation or demonstration i
of a procedural step.
8.2.4' The monitor may at any time during the analysis phase, request an explanation or demonstration of a procedural step.
8.2.5 Upon completion of the ac.alysis phase, formal analysis results I
report will be completed.
8.2.6 The formal analysis report and all acquired data will be transfered to the PDA 8.2.7 Re-looks or re-examinations may be performed as specified in the formal procedure.
1 9.0 RESULTS REPORTING The error in location and sizing of the flaws will be determined and the basis for this l
determination is provided below. VY will be responsible to determine whether the ISI procedure is adequate.
9.1 Detection Detection results will be reported as percentage of flaws detected. To receive credit for detecting a flaw, it must be reported within.50 inch ofits true axial and circumferential position 9.2 Location Accuracy The accuracy oflocating the circumferential and axial position of flaws will be reported a: rms error. RMS error is expressed as:
e#0
~
}
l w
1(True - Measured)2 l
RMS=g y
l 9.3 Depth and Length Smng
)
RMS error of flaw sizing results will be calculated. Linear regression analysis of the demonstration results may also be performed as an aid 'm evaluating performance.
10.0 RE-TESTING In the event that the results of the demonstration are determined to be inadequate to VY, re-testing will commence upon review and approval by the PDA subject to availability of mock-ups and PDA staff.
11.0 DOCUMENTATION Upon completion of the demonstration, a copy of all demonstration documentation will be retained by the performance demonstration administrator. This documentation includes as a minimum: all ultrasonic data acquired on the specimens, identification of personnel, NDE procedures, equipment identi6 cation, specimen information used during the demonstration, and the results of the performance demonstration. The tmth ' formation m
on the mock-ups will be retained by the PDA however a copy of demonstration results may also be maintained by the vendor.
12.0 SECURITY VY t d the EPRI NDE Center as PDA's will be responsible for controlling the test sperioens and maintaining the secrecy of the test specimen details. The PDA will also be responsible for maintaining the test keys during the blind demonstration and ensuring the test samples contain flaws which can be detected.
t pay 47of47
______ _ _ ____