.- . _ _ ..

. ,

,e NUR EG/CR-2149'7

[ ORNL/TM-7836 .v.

| Ag'-

. ,

N77gpb '

Eddy-Current inspection forUNION 6 eam Generator Tubing |

CARBIDE Prograrn Annual ProgressReport for Period Ending

D ecember 31,1980

C.V.DoddW. E. Deeds

*

R. W. McClung

|

|

Prepared for theU.S. Nuclear Regulatory Commission

Office of Nuclear Regulatory ResearchUnder Interagency Agreement DOE 40-551-75

.

4

SLA'!"u8i??**"[,CR-2149 R

,

L

. .

.

.

Printed in the Unitt d States of America. Available fromNational Technical information Service

U.S. Department of Commerce5285 Port Royal Road, Spnngfield, Virginia 22161

'_

Available from

GPO Sales Program

Division of Technical Information and Document ControlU.S. Nuclear Regulatory Commission

Washington, D.C. 20555.

.

This report was prepared as an account of work sponsored by an agency of theUnited States Government. Neither the U nited States Government nor any agencythereof, nor any of their employees, makes any we'isnty, empress or implied, orassumes any legal habihty or responsibi,.ty for the accuracy, completeness, orusefu! ness of any inforrnation, apparatus, product, or process disclosed, orrepresents thallts use would not infringe privately owned rights Reference nereinto any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise, does not necessarily constitute or imply itsendorsement, recomm.pndation, or favoring by the United States Government orany agency thereof The views and opinions of authors expressed herein do notnecessarily state or reflect those of theUnited States Government or any agencythereof.

.

.

_ _ _ _ _ _ _ _ . _ . _ . . . _ _ _ _ . _ . _ _ _ _ . . _ . _ .

- - _ . .

. .

NUREC/CR-2149ORNL/TM-7836Distribution

*

Category R5

.

Contract No. W-7405-eng-26

METALS AND CERAMICS DIVISION

EDDY-CURRENT INSPECTION FOR STEAM GENERATOR TUBING PROGRAM ANNUALPROGRESS REPORT FOR PERIOD ENDING DECEMBER 31, 1980

C. V. Dodd, W. E. Deeds, and R. W. McClung|

.

Manuscript Completed June 30, 1981.

Date Published - August 1981,

NOTICE: This document contains information of apreliminary nature. It is subject to revisionor correction and therefore does not represent afinal report.

Prepared for theU.S. Nuclear Regulatory Commission

Of fice of Nuclear Regulatory ResearchWashington, DC 20555

Under Interagency Agreement DOE 40-551-75NRC FIN No. B0417

OAK RIDGE NATIONAL LABORATORY,

Oak Ridge, Tennessee 37830operated by

UNION CARBIDE CORPORATION.

for theDEPARTMENT OF ENERGY

- _ - _ - _ _ _ _ _ _ - - _ _ _ _ _ _ - _ _ _ _ .

_ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ .__ _ _ _ __ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ .

J

j. .

! CONTENTS"

SUMMARY . . . . 1. ... .......... ............

INTRODUCTION 2; .. ... ...... ....... ..........

ANALYSIS OF.PROBLE . 2.. ..... ................ ,

RESULTS OF CALCU1ATIONS . . . . . . 5............. ...

EXPERIMENTAL MEASURMENTS IN THE LABORATORY . . . . . . . . - . . . . 6| *

DEVELOPMENT OF INSTRUMENTATION FOR FIELD TESTING 9.........

J PRESENT STATUS OF DEVEL0lHENT . . . . . . . 11............

) REFERENCES 12...... ...... .................

i

r

4

N

t e

i e

.

t

il

i

,

? -e

,

9

iii

_ . , . = _ _ . . . . - _ _ - _ . . _ _ _ _ _ , _ . - . . . - , _ , , _ . . . . . _ - - _ _ _ . _ . . - ~ _ _ _ _ , _ _ . . _ . . , _ _ _ , . _ _ , , _ , . , _ . _ , _ . , . . , _ _ -.

. .

EDDY-CURRENT INSPECTION FOR STEAM CENERATOR TUBING PROGRAM ANNUA 1PROGRESS REPORT FOR PERIOD ENDING DECEMBER 31, 1980

.

C. V. Dodd, W. E. Deeds, and R. W. McClung.

SUMMARY

Eddy-current methods provide the best in-service inspectionof steam generator tubing, but present techniques can produceambiguity because of the many independent variables that affectthe signals. The current levelopment program has used mathemat-ical models and developed or modified computer programs todesign optimum probes, instrumentation, and techniques formultifrequency, multiproperty examinations. Interactive calcu-lations and experimental measurements have been made with theuse of modular eddy-current instrumentation and a minicomputer.These establish tne coefficients for the complex equations thatdefine the values of the desired properties (and the attainableaccuracy) despite changes in other significant vctiables. The-

computer programs for calculating the accuracy with which var-ious properties can be measured indicate that the tubing wallthickness and the defect size can be measured much more accu-*

rately than is currently required, even when other propertiesare varying. Our experimental measurements have confirmed theseresults, although more testing is needed for all the differentcombinations of cases and different types of defects.

To facilitate the extensive labaratory scanning of thematrix of specimens that are necessary to develop algorithms fordetection and analysis for all the possible combinations ofpositions of flaws, tube supports, and probe coils, we havedesigned, constructed, and used a computer-controlled automaticpositioner. We have demonstrated the ability to overcome thelarge signals produced by the edge of the tube supports. An

advanced microcomputer has been designed, constructed, andinstalled in the instrumentation to control the examination andprovide real-time calculations of the resired properties fordisplay recording during the scanning of the tube.

We have calibrated the instrumentation on an array of stan-dard samples with all combinations of probable property varia-tions and used these measurements to formulate the algorithmsfor determining the desired properties, such as wall thicknessand defects, even when other properties are varying. We haveoutfitted a mobile laboratory, =ounted in a pickup truck, with

,

the equipment needed to make in-service inspection of steamgenerator tubing at various reactor sites, and we have success-fully inspected steam generator tubes at the Robert E. Ginna and.

Point Beach reactors.

1

_ _ _ _ _ - _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - _ _ _ - _ - _ - _ - _ _ _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ - _ _ - . _ - _ _ _ _ _ _ _ - _ - _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

_ _ _ _ - _ - _ _ _ _ _ - _ _ - _ - _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ __ _ ____ _ _ _ _ ____ ___ _ __ - _ _ - - _ _ _ _ _ _

. .

INTRODUCTION

.

The Oak Ridge National Laboratory has undertaken a program to improvethe eddy-current inspection capabilities of in-service inspection of steamgenerators for the Nuclear Regulatory Commission. .

As the interface between the primary and secondary coolant systems,the steam generator is a crucial component. Degraded tubes in the steamgenerator may increase the risk of a loss-of-coolant (LOC) accident fromthe coincident failure of several tubes either as the initial failure inan accident or as the result of pressure surges in the system resultingf rom a break and depressurization of either the primary or secondarysystem. The coincident failure of several (10 to 30) tubes would resultin a complex LOC accident that would substantially challenge both theemergency cooling system and containment isolation. The main objective ofthis program is to improve the ability to detect and determine the size oftubing defects and thereby provide greater assurance of steam generatorintegrity.

The present examination methods cannot detec.t the, cracking and someof the corrosion around dented tube supports. Low-volume flaws are alsolost at the edges of the tubesheet, at the tube supports, and sometines in

'

the bent region.Al so, accurate measurements of the wall thickness and tube inside

diameter are needed to detect continued degradation of the tubing. For,

some reactors if the average degradation rate is greater than 1% betweeninspections. a growth factor must be incorporated in the plugging limit(Reg. Guide 1.121).1 However, the present eddy-current tests cannotapproach the accuracy required for this measurement.

The three-f requency eddy-current inspection being developed by ORNLhas the potential for solving these problems. The design and developmentprogram at ORNL is following the outline listed below:

1. Calculate the instrument readings that would be obtained for alarge set of test properties.

2. Perform a least squares fit of the properties to the readings.; Determine how well the properties can be measured f rom the readings.

3. When steps 1 and 2 show the best (or adequate) accuracy, con-struct the system.

4. Measure the instrument readings for a large set of test prop-i erties. Perform a least squares fit for the properties to the measured

readings.5. Calculate the properties from the readings in real time by using

a minicomputer.6. Program a microcomputer in the instrument to perform the property

calculations as the tube is scanned.! 7. Test the instrumentation in the field.

8. Make the needed improvements on the system and retest. .

-

ANALYSIS OF PROBLEM

We have analyzed in the general way the problem shown in the blockdiagram of Fig. 1. The eddy-current probe, of whatever type, is fed with

. - - - _ _ _ _ _ - - - _ - - _ _ - _ _ _ _ - _ - _ _ - _ - _ _ -- _ - _ - _ _ _ . ._. .__

. .

3

ORNL- NG 80-19392MULTIPLE INFoRMAttoNSIGNAL

MULTIPLE FREoVENCY,MULTIPLEX FREoUENCIES, RECORDER*

oR PULSES olGITAL oR ANALOG

i b i i*

I P

I

pro 8E MULTIPLE INFoR'4ATioNVARIA8LE SEPARATOR N

DIFFERENTIAL oR A8 SOLUTE, SIGNAL DECODER REcoGNITiottCOAXIAL oR PERPENDICULAR, FREoVENCY OR COMPUTERMULTIColl ARRAYS TIME FILTERS COMPUTER'

i b

i f I P i f

DECISloN MAKING,

MANUAL oR COMPUTER

Fig. 1. Generalized inspection system.

a complex signal containing multiple amplitude or frequeacy information.The pa t being inspected interacts with the probe to modify the outputsignal, which is then decoded into the separate frequency or time

'

components, which can in turn be sent to the variable separator (andpossibly recorded). he variable separator uses the multiple pieces ofinformation, such as amplitudes and phases at various frequencies or.

amplitudes at various times in a pulse, to calculate the various proper-ties or " variables" via p grams that have been stored in a computer. Theoutput vatiables can be recorded, used for making decisions, or fed into apattern recognition system. He latter can be programmed to recognizecertain patterns of variables as benign or dangerous, and the types ofpatterns can be recorded as well as used for making decisions.

A typical steam generator tube is shown in Fig. 2, along with anabsolute eddy-current coil. he test properties that may vary in theeddy-current test are probe-to-tube wobble, tube wall thickness, the sizeand location of defects in the tube, the tube-to-support distance, and thetube support axial location along the tube. We can uniquely determinethese test properties from the instrument readings if there are at leastas many independent readings as there are property variations. We can gettwo independent readings at each frequency, and the frequencies should beat least a factor of 2 apart. Le problem therefore is to determine whichf requencies and coil designs will give us the best determination of thetest properties.

The problem can be approximated by multiple cylindrical conductors,as shown in Fig. 3. His problem has been solved to a high degree ofaccuracy.2,3 The main approximation is that the effects of the edge ofthe tube support plate cannot be calculated accurately (these are includedin the experimental measurements).-

The computer program (MULTUB) calculates the effect of the propertiesshown in Fig. 3 to a high degree of accuracy and performs a least squares

*

fitting to find the best polynomial of a user-chosen type for calculatingany desired property from the computed data. Le MULTUB program combinesthe earlier program, TUBRDG and TUBFIT, which do the calculations and pro-perty fitting separately. MULTUB has also been modified to accept largerquantities of data and to make calculations for either absolute coils orones in a bridge circuit.

- _ _, - , . _ . - -- -. . -. _ _ - - .

4 -

<

ORNL DWG 80-19391w.,

, ~'|,

A *pL)4,-

L ," !.

N'.

<?k' f 1-

DISTANCE /

.r y||DWEFECTp

Y cp ' Q E fTcy%

/s#.

Fig. 2. Test properties that vary during a steam generatorinspection.

-

L-DWG 73-4858

.. ---

. . . _ ' ',.

..

[

* ic , = >'i, '

'

.-

.

Fig. Multiple cylindrical conductors encircling and encircled by*'.

two coils in the same radial region.

- _ _ - _ _ _ _ _ _ _ _ _ - - _ _ - _ . - _ _ __ _ _ - _ _ _ - - _ _ - _ _ _ _ - _

5 -'

RESJLTS OF CALCULATIONS

The results of some of the -theoretical calculations are shown inFig. 4, where the calculated eddy-current magnitudes and phases have beenplotted for various combinations of wall thickness and lif toff for tubing.

with the nominsi resistivity of steam generator tubing and no defects,using a coil of type 83B at a frequency of 100 kHz. We see that ti. all

thickness can be determined primarily from the phase shift, but the lift-off affects both magnitude and phase. However, given both the magnitudeand phase, then both variables can be measured, provided they are onlyquantities that vary and also affect the readings. Some points labelled" Point Beach Sampl'e" have also been plotted on Fig. 4. These are thevalues of magnitW e and phase actually measured with two coils of type 838in a sample from tube 2073 140 mm (5.5 in.) below the top side of the tubesheet in the hot leg of Steam Generator A of th toint Beach reactor. We

see that the Point Beach sample measurements indicate wall thicknessesapproximately midway between 1.298 and 1.0'4 mm (0.0511 and 0.0423 in.)[ nominal thickness is 1.27 mm (0.050 in.)], and the lift-offs are approxi-mately 40 pm (0.0015 in.), indicating some nonconducting coating on themetal. Thus, the sample showed a 49- to 124-pm (3.9 to 4.9 mil) decreasein wall thickness, with a 40 pm thick (1.5-mil) nonconductive coating.-

The sample has had definite change in its electrical and magneticproperties.

.

ORNL-DWG 80-121666.0

%

LIf

POINT**

BEACH.

+fSAMPLE [

> 5.0 m

1.298-mm p- WALL 1,074.mma

H WALL %b 0.914e m

.L_lFT,0cF 0,ya mm$ ,

M~~

- _ _ _ \_

4.0 '% LIFT. Opp g ,9

'.

1.0 1.1 1.2 1.3 1.4,

PHASE (deg)

Fig. 4. Magnitudes and phases for various lift-offs and wallthicknesses.

. _ _ - _ _ _ - - _ - - _ - - - - - _ _ _ _ _ _ _ - _ _ - _ - _ _ _ _ . - _ - _ _ _ . - - _ _ _ _ - _ - _ _ - _ - _ - _ _ _ - _ _ _ _ _ _ _ - - _ _ _ _ -__ ___ _ _ _-_ _ _ _ _ _ _ _ _ _ - _ _ -

6

We also measured the wall thickness using a through-transmission

measurement. We tube measured 1.16 to 1.20 un (0.0457 to 0.0472 in.),

with an average value of 1.18 mm (0.0464 in.). h e average reading from

the previous measurements with a reflection coil was 1.19 mm (0.0467 in.).We will loan a through-transmission instrument to Westinghouse to test itsother tubing samples. Westinghouse has not been making absolute wallthickness measurements on the samples in its laboratory. There is some -

speculation that the pulling operation may ha n caused the decrease inwall thickness.

EXPERIMENTAL MEASUREMENTS IN THE LABORATORY

Experimental measurements were performed m steam generator tubingsamples to verify the analytical results and to include the tube supportand tubesheet edge effects, which the theory neglected. Also, second-

order effects such as variations in some of the coil and cable construc-tion details are included in the measurements.

A block diagram of the three-frequency instrument is shown in Fig. 5.The instrument consists of three separate oscillators from which the

-ORNL-DWG 78-1618R

,

M AG t !*

BANDPASSMP t

OSCILLATOR t -

PHASE- * DETECTOR +

iI e! DEFECT SIZE _

, MuER! '

OSClLLATOR 2 ' ;; AND :

d PCwER AMP' OEFECTBANDP'.SS ,4,

AMP 2 LOCATION

DE 00 A NG( COtt TO TUBEOSCILLATOR 3 m

0 STANCEgPHASEp

n -+ DETECTOR +-2 TUBE TO SUPPORT

DISTANCE[ PHASE 2--

|' TUBE WALL

+ THICKNESSgyp 3

|*

PHASE* DETECTOP +

| 5

| PHASE 3 ,

u_

Fig. 5. Block diagram of a three-f requency instrument.

.

7 ,

lsignals are mixed before the composite signal is transmitted to the probe |through a power amplifier. The signal from the probe is separated back |into three discrete frequencies.by using bandpass amplifiers. The magni- '

. tude and phase of each signal are then sent to a demodulating computer,which digitizes the readings and calculates the different properties.

The demodulating computer in our development program can be either theMODCOMP IV minicomputer or the NDT-COMP 9A microcomputer that is in the |

-

instrument. The three-frequency instrument 4 that has been developed at '

ORNL is shown in Fig. 6. A few of the modifications and improv-ements to,

the instrumentation package have been discussed in the three pteceding i

qua rte rlies. 5-7 These include the antomatic calibration module, the !magnetic tape drive, the COMP 9A microcomputer, and the aaalog-to-digital jconverter module. The instrument has a calibrator module that can becontrolled by either the microcomputer or the minicomputer. The calibra-tor will switch in passive R-L-C networks that will produce a known magni-tude and phase in the instrument. The computer than records thesereadings and does a least squares correction of all subsequent readings.

|,

;OL W PHOTO 1992-80

.

/W

n-

t ..

bff -

-** g$&

- t. ~ w . M h)[(- a> gq m., w w ,._

- .. y *g.

..

, g .

| [' 'L' yb , i f' -

,' ? q _ n,

.. -~.-% Yt- , A& 25 4& v

_

" ^ - m'- 1 ; r a, b _ s ,[_,,, _

-' '

'

_

Fig. 6. The three-freqitency eddy current instrument.

The NDT-COMP 9A microcomputer,8 which is used to control the instrument in,

the field, is shown in Fig. 7, in the computer module of the instrument. 1

The microcomputer can have as much as 16K bytes of memory, divided between| PROM and RAM, and it has a floating point hardware chip. The microcom- |

puter is 8080 based and has 72 parallel input-output ports, one serial i.

port, and three counter chips.;

The MODCOMP IV computer is used to run the test while the instrument !is in the laboratory because of the ease in programming the MODCOMP IV,-

its larger data taking capacity, and its faster operating speed. For theexperimental measurements, as well as the analytical calculations, a large

|i

1

-.-- - -_ - _.__ .-__- -- . - -...-.. - - -..- - _ . _ - - _ - . =

.._. . _. . . _ _ __ _ __ _ _ _ - _ _ _. -_. . _ _ _ - . _ _ _ _ _ . _

8-

|

* *:4h [[;ff~~"VNWg ; P ' ' f G * * ~

. ou ;nsTogyf.}7. + < q. , _ ,- - s . ,, y. ,

M. ,

?,' -g . 2 .

h. 2,, .a .

i 'f w &

by.k .,

~

l'- -;5e ..

g,:, , . 4% .

.s._ ,.-g ' ' ;

.(-

,,

y n .,e r - .,

'. .! l'

,% - b $ *.

;< ;. %'

..

- .- ;% - ( ',,

'* ~. .s.. ,

\}~'g> 4 ..

W{44 . k -| .N w

. .

. % ,# ' r ' :- +D ., ',ff . ;' [-.,

, * ' ;4'

; /. 2 ,

~, g. ,3

a_ .s. .-.

x ,_ - ,

%,,g, % 7 i'

,'p -N ;, - p .,j

-

.. q

^ R;x. - 3;ifj, f q <, %,.'Q

->.' i, % Q3 ,

4

. 9 lj p-;/%.,? q j6% n -h ,

.. . , ~ n .- ... g,% s - '.-

'

f ,; Nyip .,-

'

.

., -. s .- < .(n

Fig. 7. fhe NDT-COMP 9A microcomputer in the three-f requency instru-ment computer module.

,

{number of different properties are considered. For instance, if an exper-

i iment has three variations of tube wall thicknesses, three variations intube inner diameters, two different flaw sizes, two different flaw loca-'

tions, three variations in tube support diameters, and ten different loca-tions of the tube support along the tube (e.g., relative to flaws, probe,etc.), there are 2160 different combinations for these parameters.|

iBecause of the large number of possible variations of the tube and tube-support positions with respect to the coil, the accurate positioning forall these cases without making any mistakes is very difficult. 'Ihe re f o re , |we have constructed an automatic positioner, which is controlled by the

IMODCOM2 IV and is accurate and repeatabic to within 10.05 mn (0.002 in.).A new tu!:e standard with 5 thicknesses, 3 inside diameters, and

:20 different hole sizes has been constructed. It has been run withTUBRDG, TUBFIT, and PLTRDC. 'Ihe runs thus far have shewn some improve-

ment, and we are arranging our properties in different manners to get ,

additicaal improvements.

;I

e-,- ,,c- _w--r-.----.- _- - - ~ - - - - - - - - - - , - -

9. .

The fit results are somewhat better than the previous standard, andeven better fits can be obtained if we break the tube into regions alongthe length. Our computer simulations can predict the instrument response

' when the tube is free or when it is completely inside the tubesheet ortube support region, but not when the probe is in the interface region.

as wellHowever, experimental measurements are valid in the latter region,as the others. Therefore, we have arranged the program so that the com-,

puter can determine if the probe is in the free tube, the tubesheet, oran interface between the tube and tubesheet and then pick the proper setof coefficients to give the best fit. While all property calculations areimproved by this technique, the improvement is,necessary only for thedefect calculations. This. concept will be-tested first on the MODCOMPand, if successful, programmed for the microcomputer.

DEVELOIMENT OF INSTRUMENTATION FOR FIELD TESTING

We have developed remote controls for the instrumentation. Only

the eddy-current instrument and the probe pusher-puller need to be placedinside the containment, with the controls and the recording equipment

.

taside a van about 120 m (400 ft) from the instrument, as shown in Fig. 8.

ORNL-DWG 80-12167R*

INSIDE CONTAINMENT

I Oi

STEAM" 'T

Fig. 8. Eddy-CurrentInspection of Steam GeneratorConducted by Operators Outside pContainment.

i

'

PROBE yPOSITIONER #

NPROBECONTROL AND DATA BUSS -

DRIVEOPERATOR AND-

DATA STATION _

b\ EDDY-CU R RE NT

-

-

gINST RUMENT,

- . - . , . _ - . . - - - - . - . - ..- . - - .

,_. _ .__ __ __ - __ _ ..____ _

!

| 10! -

l

iThe remote operation will allow us to use trained operators at our remotedata station for the entire inspection without fear of their becoming

| " burned out."

We first tested the instrument wh1_e it was contained in a plastic ,

bag, as would be required for the inspection inside a centaminated zone.The instrument worked fine, except for a minor problem of the terminalreceive chip overheating during troubleshooting. (In normal operation for -

steam generator examinations, this chip is not used.) This problem was'

cured by preselecting a chip (apparently identical) that does notoverheat.

A variable speed selector has been programmed for and tested with themotor drive.

A trip was made to the Point Beach Reactor (under separate funding)and the eddy-current inspection using present commercial equipment wasreviewed.

The truck containing the control equipment is shown in Fig. 9, andFig. 10 shows a view of the interior of the van, with an opcrator at the

; eWLew "r !ORNL-PHOTO-5752-80

-

: .. ;- -

( '~ Z C a 2 f dnn~~. _..%.,

-_

,

. .-

Fig. 9. Mobile eddy-current inspection laboratory during checkout ofremote probe operation. Note the cable from the right rear of the truckleading into the building.

computer terminal that communicates with the microcomp r in the eddy-current ins t rum ent. The operator can position the probs at a given steamgenerator tube with a Westinghouse positioner and the closed-circuittelevision monitor shown in the upper right of the picture. Data arerecorded on a strip chart recorder in the lower left and a digital tape3

recorder in the lower right of the picture. -,

t

|

_ _ __ _ _ _ _ _ - - . - - - . _ _ - _ _ _ _ _ _ _ _ _ _

11.

, - - - - - m .5, ORNL-PHOTO-5753-80. . . - . m ,

..g6..t, Y]I J q 3'

'

!,. . (_-.ym,

w,-v :: .a . .

~ + , ,,

! i1 * .,

~ ''$

t'*; ,,

ll f' f, f .)' 5# 11\ f '-

.

y. s;I :6 y }. M - e

.

!} g - -

*

7D

.WW. . . . .

, a_, ; -f

n

,.

.

I'

| L- L L i m ,

Fig. 10. Operator at remote inspection station in the truck.

PRESENT STATUS OF DEVELOPMENT

!

We took our instrument truck and related equipment to theRobert E. Ginna Nuclear Power Plant near Rochester, New York, inNovember 1980. The equipment functioned well and we recorded good dataon the magnetic tapes. There was an offset drif t of all the readings onthe thickness channel, including the readings on the standard, of about

0.30 mm (-0.012 in). This drif t was due to the large temperature dif-ference between the laboratory and field. A calibration reading is madeon a standard in the laboratory and this is repeated in the field. Themagnitude and phase readings at different frequencies in the field arecompared to those in the laboratory, and a correction is calculated tomake tue readings the same. This correction eliminates the temperaturedrif t of the cable and is applied along with the correction factor thateliminates the instrument temperature drift. The of fset in thickness at

Ginna was corrected in the processiag of the magnetic ttpes and aprogramming change was made in the microcomputer to auttnatically correctthe drift before the Point Beach trip. The microcomputer was also

programmed to automatically add the end-of-file (EOF) m -k t( the magnetict v- inis is used to mark the end of the data for each tube and allows

.

- - - + - . - + - - - - __..e-,,,,,,-,-___.m,ym.,,,%-%- .__,.,,__.-,e.y.,,,,,,g, .

12-*

high-speed access of each tube's data. We forgot to write an EOF af terone scan at Ginna, and the data from two tubes were run together. Thescan of the PNL laboratory sample with intergranular attack (IGA) of 5 mil .

(40 p) showed a wall thinning of about 5 mil (40 p) on the wall thicknesschannel. There were small decreases on the order of 1 to 5 mil (25 to40 p) in the wall thickness in the same tube that Ginna personnel *

suspected of IGA.

In December 1980 we performed a similar inspection of the Point BeachUnit 1 Reactor at Two Creeks, Wisconsin. The frequency had been changedfrom 10, 100, and 1000 kHz to 20,100, and 500 kHz to improve thestability. The wall thickness varied considerably in the region above thetubesheet and appeared to be thicker than 1.27 mm (0.050 in). The tubesgave the correct thickness in the tubesheet region except that a muchlarger proportion of the tubes showed a slight decrease in wall thickness,which may be attributed to IGA.

The microcomputer in the instrument computed the wall thickness andtube inside diameter in real time and plotted them on a stripchartrecorder. We had intended for the person running the probe in and out ofthe tube to also watch the data and report any abnormalities as each tubewas scanned but it proved to be too much for one person to do. We there-f ore recommend that one person manipulate the probe and another observethe data. '

The raw data and calibrations were recorded on magnetic tape. If

property variations not anticipated (such as the Point Beach readings .

above the tubesheet) occur, the system can be retrained for theseparticular properties and th'en the properties recalculated from the rawdata on the tape using the new coefficients. Likewise, if metallographicexamination of the tube samples shows that IGA produced in the laboratory(the PNL sample) is not the same as IGA produced in the tubesheet region,the system can use the sectioned tubes as standards and rerun the tapesfro, both Ginna and Point Beach.

Thus the system has demonstrated the ability to calculate the desiredproperties in real time and to allow the analysis or reanalysis of addi-tional properties back in the laboratory. Various causes for the apparentwall thickness above the tubesheet will be investigated.

REFERENCES

1. U.S. Nuclear Regulatory Commission, Bases for Plugging Degraded IFTSteam Cenerator Thbes, Regulatory Guide 1.121 (August 1975?-

2. C. V. Dodd, C. C. Cheng, and W. E. Deeds, " Induction Coils coaxialwith an Artitrary Number of Cylindrical Conductors," J. Appl. Phys.45(2): 638-47 (February 1974).,

3. W. E. Deeds and C. V. Dodd, Multiple Property Variations in CoaxialCylindrical Conductors Determined uith Multiple-Frequency EddyCurrents, ORNL/NUREG/TM-335, NUREG/CR-0967 (November 1979).

4. C. V. Dodd and L. D. Chitwood, Three-Frequency Eddy-Current instrwnent "

for Multiple Property Problems, ORNL-5495 (March 1979).

.

, + , . . - - . ~ . - - . - , . - -. , . . . - , , - , .-., -.,.

. - _ - - - - =. - . .- . . -

l'). .

5. C. V. Dodd, W. E. Deeds, and R. ,W. McClung, Eddy-Current Inspection;

for Steam Genembor Tubing Prognm Quart. Prog. Rep. Mir. 31, 1979,ORNL/NUREC/TM-341, NUREG/CR-0918.

6. Ibid. June 30,1979, ORNL/NUREG/TM-353, NUREG/CR-1069.-,

7. Ibid. Sept. 30,1979, ORNL/NUREG/TM-377, NUREG/CR-1372.8. C. V. Dodd and R. F. Cowan, The NDT-COMPB Nicrocoriputer,4

*; ORNL/NUREG/TM-390, NUREC/CR-1548 (September 1980).

.

t

9

i

t.

I!

i-

t

O

_

*

_ . . - _ . _ . . _ _ _ _ - . - __ -,-_-.---_ -- --.__._.---_- __--.___ - - _-_--.__- --__- - _ - ____ - - - - . . _ - - -ar-_a1 .__a_w

_ _ _ __ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ . _

. .

15

NUREG/CR-2149ORNL/TM-7836Di s tribution-

Category R5'

INTERNAL DISTRIBUTION

'

1-2. Central Research Library 23. J. M. Slaughter3. Document Reference Section 24. J. H. Suiith

4-5. Laboratory Records Department 25. B. G. Eads6. Iaboratory Records, ORNL RC 26. W. Fulkerso..7. ORNL Patent Section 27-29. M. R. Hill8. Nuclear Safety Information Center 30. J. M. Jansen, Jr.9. C. R. Brinkman 31-35. R. W. McClung

10. F. N. Case 36. F. R. Mynatt11. W. R. Casto 37. A. R. Olsen12. J. A. Cox 38. D. C. Parzyck

13-17. W. E. Deeds 39. G. W. Scott18-22. C. V. Dodd

4

EXTERNAL DISTRIBUTION.

40. NRC, OFFICE OF NUCLEAR REGULATORY RESEARCH, Washington, DC 20555Program Sponsor

41. DOE, OAK RIDGE OPERATIONS OFFICE, P.O. Box E, Oak Ridge, TN 37830

Of fice of Assistant Manage r for Energy Research and Development

4243. DOE, TECHNICAL INFORMATION CENTER, P.O. Box 62, Oak Ridge, TN 37830

44-433. For distribution category R5 (10 - NTIS)

.

.

*U.S. GOVERNMENT PRINTING OFFICE: 1961 4 40-062/155

,..

. .

. . _ . _ __

"nc renu 335~~

-

"" " " 'n,m U.S. NUCLE AR REGUL ATORY CCMWSSION ' " .*

NUREG/CR-2149BlDLIOGRAPHIC DATA SHEET ORNL/TM-7836

4. TITLE AND SU8 TITLE (Add Vo/ums Na. i/ appicpr,,ref 2. (Leave blan Al

Eddy-Current Inspection for Steam GeneratorTubing Program Annual Progress Report for 3. RECIPIENT'S ACCESSION NO.

Period Ending December 31, 19807. AUTHORG)

. C.V. Dodd W.E. Deeds u On m | naa5. D ATE REPORT COMPLETED \

R.W. McClung Auaust 14R19. PERFORMING ORGANIZATION NAME AND MAILt IG ADDRESS (Include Zip Code) DATE REPORT ISSUED

Oak Ridge National Laboratory "A"uEust I "N81Oak Ridge, TN 37830

6. (Leave blank)

8. (Leave Nank}

12. SPONSORING ORGANIZATION NAME AND MAILING ADDRESS (Include Zip Codel10. PROJE CT/ TASK / WORK UNIT NO.

Office of Naclear Regulatory ResearchU.S. Nuclear Regulatory Commission '11. CONTRACT NO.Washington, DC 20555

FIN No. 80147

13. TYPE OF REPORT PE RIOD COVE RE D (Inclusive dares)

15. SUPPLEMENTARY NOTES 14. (Leave Naakl |,

16. ABSTR ACT 000 words or ;ess)

Eddy-current methods provide the best in-service inspection of steam generator tubing,1 but present techniques can produce ambiguity because of the many independent varia-

bles that affect the signals. The current development program has used mathematical'

models and developed or modified computer programs to design optimum probes, instru-mentation, and techniques for multifrequency, multiproperty examinations. Interac-tive calculations and experimental measurements have been made with the use of modulareddy-current instrumentation and a minicomputer. These establish the coefficients forthe complex equations that define the~ values of the desired properties (and the attain-able accuracy) despite changes in other significant variables. The computer programsfor calculating the accuracy with which various properties can be measured indicatethat the tubing wall thickness and the defect size can be measured much more accu-rately than is currently required, even when other properties ar<3 varying. Our experi-mental measurements have confirmed these results, although more testing is needed forall the different combinations of cases and different types of defects.

17. KEY,WORDS AND DOCUMENT ANALYSIS 17a. DESCRIPTORS

.

~'.

-y

17tx IDENTIFIERS /OPEN ENDED TERM.CW 5ss

21. NO. OF PAGES19. 5b-CUlITY CLASS /Thss reporrl'8. AVAILABILITY STATEMENT

nc assifiedS Unlimited~ L 'M 22 PnsCE20.S99y gn.so ort+-

- , - . . , , . . . . . ,__ _ _ _ ._. . . _ _ _