Skip to main content

Full text of "Stress corrosion crack detection in alloy 600 in high temperature caustic."

See other formats


“Calhoun 


Institutional Archive of the Naval Postgraduate School 





Calhoun: The NPS Institutional Archive 
DSpace Repository 


Theses and Dissertations l. Thesis and Dissertation Collection, all items 


1996 


Stress corrosion crack detection in alloy 600 
in high temperature caustic. 


Brisson, Bruce W. 


Monterey California. Naval Postgraduate School 
http://hdl.handle.net/10945/9017 


This publication is a work of the U.S. Government as defined in Title 17, United 
States Code, Section 101. Copyright protection is not available for this work in the 
United States. 


Downloaded from NPS Archive: Calhoun 


Calhoun is the Naval Postgraduate School's public access digital repository for 
| г D U DLE Y research materials and institutional publications created by the NPS community. 
FW узу, Calhoun is named for Professor of Mathematics Guy K. Calhoun, NPS's first 


ІШ KNOX appointed — and published — scholarly author. 

OM LIBRARY Dudley Knox Library / Naval Postgraduate School 

411 Dyer Road / 1 University Circle 
Monterey, California USA 93943 





http://www.nps.edu/library 


n A de 
HO rA EC 
MUI V TM 

АЕК] 


у Ji 
one 
ne NO i D 
АҒУЫ, ЕР 
En Г НУНУН 
К MUN 
Ё jj iter М 
(қ 
Hn 


Fus) 


DR 
IPOD TOI 
E 
am р 
ком би 
ШАГ 


o E 
IEEE 
ЫҢ БҮНЫ) } 


~ ` 
UA 
m И 
И 
M А d" EA 
pas: р 4 EAS ы А A МЕ Ta 
TIT 7 жі EDEN р 
ЗЕН МИ: Ро 
и a CREE 
is ЕВЕ Хү te " А e KRAFT 
ER OR Seat E А, PHI MERA М 
ІЛІК H К AE M ун 
А БА Итин Е 
Mi УТИ w 


à 


АИЙ НЕ 


"nn 
mu DISCE 


ЖОО» а 


4 
Em 
" 


H 
СРОДНИ 
DET o ЖАҒАСЫН 


mn 
[ж] ПС 
Ne 


ОГ 
] 
TR 
A ER 
NEP ee 
" ШЕТІ 
ТЕГ КІСІНЕП 
E ae te | 
B КЕРУУ 
] noo 
4 две у деле 


mr 
"аз 
П 


ig: 


147.” 
ne 


.. 
О 


САМОА ЈЕЈЛ 


ey 
CCP 
DE XH 


ud 


E 
ПИСТИ 
а 

D] 

LEAD 

te o crit 

Dum 


LT 
СЕЛУ 


И 
АЗ 

DORT NEUE, 

IDE 


RIT us 
A ПРЕ 
NETS 
TT 


ИШЕ 
КАТИ 


% t 

4 САЈМУ 

П Denm 
УАР АН 


" 
HO DE: 
m 


ЛЫМЫ ҒЫМ 
ELS Am 5% 
РОДИО А АТА 

є 


М 
ИРИ" 
DELL Bert: 


D 

МАУ 

5» LEY MUT 
ду 


ПИ 
22174 " 
LS 
EIL 
$m. nnm 
Sen da. en Pe 
ВА 


LEM 


cu 
EE 


he oo 
EE P 
P 
ARAS 
„> 


4 


Pr A, \ 
PS vi 


E SH 
"unn 


47 
КАНУНИ 
D X Hats Pn 
42545465 
vg 


3 
v 
44 
фе" 


p 
vu 
D 
4 


[4 
М чт \ 
ПОЕН 
РА АНА ЈЕ 
Ж БАЛУ 
ИЕН 
АЕ и. ВРА 
DEE 


тала 


(ЖУЛИК 
П 

г 
x 


v пей 
Улы ау 
И 


mt 4% 


.. 
EIL 
PN FEDES 


АИК 
n a И 


б 
ЕВИ 
s je 


E 
"тула 
Ua 
m 
En e wr 
меч 
РЕЙ 


COR] 
AR. 


E 
DA AN 
n ГАЛ 
DUET T. NIE Mechs FD 
Но Н 


САН 


Core 
mov 


T 


DAY 
С) 


Ubi 
| 


HEN 


n 


t 

4 

ТАЙДЫ АЗ 

ПИ 

МЫ 
КАЖ 


(o 


im А} nr 


Md 
«7% 


EN ern 


r 
А, 


үз 


et 


DR 


E Hi 


7 


n 


[ri 


Г NR 
^i Nu 
d 


KA 
Y: 


"Maid" V yn 
у 
En 


à 
ES ЫДА, 


PEDES 


[rd On NL 
NUI 


VC SHE EM AE 


is 
О 





T 


PINE 
ве 


ои ДЬ 
арра. 
ENTE: 
b 


"PEDE 
MID NT od 
274 d 
Pa 


БАЙ 
ғары 
LES 


ПЕП 
PN Т 

4 Б>; ТА 

еы br өлде PAA 
ri Hep Eb РЫ Я 
Ici ҚСТАН ҚЫСА ІІ; 
CATA RT EY: 
pte dr pe 


IE 


Ў, 
reso 


n 
De а x ur 


Fi 
ER 
P 


rc 
АТТАН ЫН 
RE re NE 
LED d a. Md 
TE 
ifr 


Y 


НИ, 


E 

мі. ee > 
622 VUA 
и: 


ts 
d Ж ru aes d. 
E rar 
Bee eat tg 
AERIS EN 
ЕЛ 


921% 
Ды LC ELE 
Е 
e 3 


DUDL. sHARY 
NAVAL POSTGRADUATE SCHOOL 
MONTEREY CA 93943-5101 











Stress Corrosion Crack Detection in Alloy 600 
in High Temperature Caustic 
by 
Bruce W. Brisson 
B.S. Mechanical Engineering 


Syracuse University 


SUBMITTED TO THE DEPARTMENT OF NUCLEAR ENGINEERING IN PARTIAL 
AUEFICEMENTOF THE REQUIREMENTS FOR THE DEGREE OF 


NUCLEAR ENGINEER 
and 
MASTER OF SCIENCE IN NUCLEAR ENGINEERING 
AT THE 
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 


JUNE 1996 
© 1996 Bruce W. Brisson. All Rights Reserved. 


The author hereby grants to MIT permission to reproduce and to distribute publicly paper and 
electronic copies of this thesis document in whole or in part. 





nil \t cy WMV i А ғ Ty 
- PA ғ LIBRAR 


ИН АКЕП SCHOOL 
MUNICMEY CA 93943-5101 


Prrees Corrosion Crack Detection in Alloy 600 In High 
Temperature Caustic 


by 
Bruce William Brisson 


submitted to the Department of Nuclear Engineering 
on May 1996 in partial fulfillment of the 

Requirements for the Degree of Nuclear Engineer and 
Master of Science in Nuclear Engineering 


Abstract 


Alloy 600, the material used for pressurized water reactor 
steam generator tubing, is susceptible to environmentally 
assisted stress corrosion cracking. Intergranular stress 
corrosion cracking (IGSCC) attacks the tubes in areas of high 
residual stress, and in crevice regions. No method has been 
successfully developed to monitor steam generator tubing in- 
situ for crack initiation and growth. Essentially all available 
published IGSCC crack growth data for alloy 600 is based on 
non-tubing material. Although it is very likely that the 
current data base is applicable to tubing processing, 
differences between tube and other geometries make a comparison 
between tubing and other data important for verification 
purposes. However, obtaining crack initiation and growth data 
from tubing is difficult due to the geometry and the thin wall 
thickness. 


In this research the goal was to develop a system to monitor 
crack initiation and growth under in-situ steam generator 
conditions in tubing and, having developed such a system, to 
then obtain data on actual alloy 600 tubing for comparison to 
published results for non-tubing material. 


A nickel autoclave system capable of testing pressurized tube 
samples was constructed. The system is capable of operating 
in either recirculating or static mode as assembled. The 
operating environment was 10% NaOH (with 0.1% Na,CO, added) 

with a hydrogen over-pressure. The test sample was internally 
pressurized and electrically isolated from the rest of the 
system. Crack initiation and growth was monitored using an 
alternating current potential drop (ACPD) method especially 
developed for this purpose. Sample polarization to +150mV with 
respect to nickel was imposed to accelerate crack initiation. 





Mill annealed alloy 600 tubing was chosen as the test material. 
A special heat of material, fabricated for industry-wide 
research by the Electric Power Research Institute was used. 


The material, heat 96834, was mill annealed at 927+14°C. 


The program was divided into two phases: (1) system/methods 
development, and (2) crack growth measurement. 


A multi-frequency ACPD system was developed for the detection 
of initiation and propagation of short cracks in alloy 600 
tubing in high temperature caustic environments typical of an 
operating steam generator tube to tube sheet crevice. The 
system was then used to monitor crack initiation and growth in 
alloy 600 tubing. 


Limited IGSCC crack growth data for the tube samples was 
obtained. Average values of K; ranged from 4.2 to 17,7 МРаүт 
ПИШЕ growth rates from 1.6 to 12.3 mm/yr. 


The data obtained compare well with other, non-tubing, data in 
the sense that there is overlap between the data from this 
program and the general body of data in the literature. 
However, the data developed here indicates (limited as the data 
is) that a power law dependence of crack growth rate on K may 
be appropriate. The general body of literature data is too 
scattered to show any such dependence. 


Thesis Supervisor: Ronald Ballinger 
Title: Associate Professor of Nuclear Engineering and Materials 
Science and Engineering 





Acknowledgments 


MEcSunotesutticiently express my gratitude for the material 
support and advice given by numerous companies and individuals in 
constructing and operating the equipment for this thesis. I 
would especially like to thank the Electric Power Research 
Institute and Dr. McIlree for their generous financial support 
and advice, Peter Stahle, our lab engineer, for his expertise and 
guidance in the many disciplines it required to make things go, 
and Dr. Hui Au for his technical expertise in software 
development. 


I extend my sincere thanks and gratitude to Professor Ronald 
Ballinger, my thesis advisor, for his guidance and persistent 
encouragement that “things really would work”. Without his 
expertise and constant remedies to apparent dead ends, this 
thesis would not have been possible. 


Finally, I would like to dedicate this work to my wife Mary 
and my father, Marcel: 


To Mary, I owe an unpayable debt for her constant 
encouragement, endless hours proof reading my papers and thesis, 
and somehow managing our four small children when I needed to 
study, write or stay at the lab late. I will never forget the 
love and support. 


To my father, Marcel, who encouraged me from first grade on 
and was always there when I needed him, but passed away before I 
started at MIT, Thanks Dad. 





Table of Contents 


A ПИ EE. sacs aee cM s s ee ee 2 
O ее. 4 
TEA O AS NA ЕТИ aste veteres e iere eie e 5 
Е ее. 7 
В... 10 
ее... 11 
о. 14 
2.1 Nuclear Steam Generator Materials and Environment ........... 14 
EE mcam ene raros (COnstruetioln КОО ООС КО ое. 14 
PEE cSrsameCenerator Chemistty..... v... o isa i nbn URL RU. ROS ee 16 
2.1.3 Material Properties of Alloy 600 Steam Generator Tubes...18 

MEM Iss Corrosion Cracking Theories .....-:4.9..- e ae e eins 22 
2 Rupture theory Lor Crack Initiation... sten 22 
Ир 2erackeEOpasation bycAnodiesDissolution..-..:. 7.2: 4.99245 26 
Naci Emopagation by Мчетотота Formation... .... <<... 28 
2.2.4 Impurities in Secondary Steam Generator Watef ............ 30 

Е о ЕЕ ea a a ne 32 
DcraelMeasurement bwePetential Drop Methode .......2- 9*5 34 
Is Ey o ЮЕ Росава Drop а. ое ре у. 34 
D ору ACUPorterntaedl Drop. x EE. rixa T RET TS 36 
DNCEHRKocentehResedg5Gi ve шшш инт ез. ш RH ео ео оа еее ES 43 
DENMUDOGEISESRMISISUEATCAPDPARATUSS C. eee ee spei us ees i s 45 
эм есап са бус сеп ресс роп ш 4 m mem e sue than nn 45 
В Еос Гамевава Heating System ааа 47 
о теи Ка оп Боор ао ена a 48 
ЕО ак пе Е Е еа ое аре а 49 
аа зашрге Pressumiodtion System... в 50 
СОПОЛ еШ e e ee A vie ЭЯ 
оа ее оу згеше A e 52 
м Е... ее 54 





a 54 


Е ВЕЕР ВОО. 57 

Seo Alternating Current Potential Drop (ACPD) System ............ 58 
E FERIMENTAL PROCEDURE een een a 61 
DN cDDESSLOOI otSbréinedbple Testing o. uer] Ee e m m 61 
астас па сТатасп апа Скомсп Меавпкелепеа .2.................. 68 
Le esa cree crete resto Mu EE I. esie sae ees ste s s d s 70 
EE С БИО ког Рштелрје тезе то fot. weds een 70 
Em Crack initiation and бкомсп Меавикетпепез .................... 78 
Som BBG (sO) Nimes ice USE Uu uS reete eum e Vor M ere re s 88 
ООО Е Erunciple а 88 
саске Бабе ee ee 0% 91 
DEXCONCLUSIONS ...:-9 92s Peer ate tet S S T S i re 95 
ее... 96 
КЕЕ ЕМС ЕЗИ О ТАТО ао о о оаа а xS 97 
БТ SYSTEM OPERATING CHARACTERISTICS re... en seen 101 
A-1-1 Omega CN9000A Heater Controller Parameter Settings ....... 101 
A TS RET QUE Ius eir rur eee ws IOS 
Ae Aoc lavemHceadiSea MRIN Farlutes ... none... osos 104 
ОН НЕЕ ЕКО a A e шз Rr e ens 106 
а Eten start-zup Procedure = Static Operation ...... 0.000. 107 
Pad 20 Stem etsbt upMDPOcedure Recirculation. een e e eea 110 
A=2-3 System Shutdown Procedure - Static Орегатісп ............. 113 
522 Sowsbem Shutdown Procedure — Recibculatqon .3.59 4.9 «ds 114 
2122-59 Emeneeneyschürdewnzbrocedurese An eres «Sale = eR rss s IIS 
А-2-6 Башр е Реевешк зар ош System Operation ана 116 
2-7 Па ено оп эуеш реба оа ао он То 
2-9 lec rro les ste kei latina Technique a није ........ 119 
AUS DE RARUE De Ss UMM DRAW LNG GS 1. 9999999533. 2 a a ei 121 


A-4. AXIAL FATIGUE PRE-CRACK SAMPLE PREPARATION PROCEDURE . 142 


A Dr ВЕРОНА e Оаа аана 146 





List of Figures 


ДЕД 


2-2 


2-4 
25 


Beh 
325 
326 


=з 
379 
4-1 
JU 


TYPICAL AREAS OF STEAM GENERATOR TUBE FAILURE BY STRESS 
CORIO Si ON ARA stele ten es uL Ure doe a bes о ES 


EVANS DIAGRAM FOR ALLOY 600 SHOWING THE EFFECT OF 
CHROMIUM CONCENTRATION ON CORROSION CURRENT DENSITY AND 


PASS TVA E nee RN s п о. 20 
SCHEMATIC OF A STEAM GENERATOR TUBE SHOWING TYPICAL 

ЛЕРД О ОЕШ ЕТ СНЕ ЕСЕ ДШ ОЕК СС o a P e v ew eoo go s 25 
ШЕР Wem CNN Cal Ole ee een 38 


CURRENT DENSITY AS A FUNCTION OF DISTANCE FROM THE 
EXTERIOR WALL OF A TUBULAR CONDUCTOR FOR VARIOUS VALUES 


Оо sien cue ont gai eee ek este clase VETERE ee ee ree 39 
Ос СОТ октор BY PICK-UP WIRES ON THE ТОВЕ a a 41 
AUTOCLAVE SYSTEM SHOWING RECIRCULATION, STATIC AND 

DISCOUNT ION VG DEMS es ee ee ee оО 
AUTOCLAVE WITH HEATING MANTLE SHOWING RECIRCULATION AND 

STANI CRCONNECTIONS E T T mms 47 
Уа РОВ СО СТВ ООН А ее S 53 
EDX OF ALLOY 600 HEAT 96834 (Low TEMPERATURE ANNEAL) ........ 56 
THREE SDIMENSTONAL SIM MICROGRAPH OF HEAT 96334 „un scene 56 
HEAT 96834 SEM MICROGRAPH SHOWING INTRAGRANULAR 

БЕК ITA TES MM re ere a a ao aa V e DU 
SIMPLIFIED SCHEMATIC OF TEST SAMPLES ILLUSTRATING 

MACHINING DIBPERENCES [Ni THE ОБЕ SECTIONS ... . 3 2.44 .60.06. O 
ТИЕ ЕРТ ОТАМЕЗ 225,05. о, e cae Caster a. In 59 
А РЮ® © МС ТЕМ ОСНЕМАЛТ СИУ (a ен ео о. 60 
ПАРАНА БӨЛ ЗАЛИВЕ О Ое кта ва ае 61 
SCHEMATIC OF SAMPLE TUBE INSTALLED INTO AUTOCLAVE HEAD ......... 63 


SAMPLE ASSEMBLED INTO AUTOCLAVE HEAD SHOWING ACPD INPUT 
AND OUTPUT WIRES, THERMOCOUPLE, LEVEL SENSING AND 


ЕЕЕ REFERENCE E LECTRODE: „ы... a. 64 
POLARIZATION CURVE FOR ALLOY 600 AT 300°C In 10% NAOH........ 67 
POTENTIAL DROP FOR THE TEST AND REFERENCE ÁREAS ............... ү 


SAMPLE TEMPERATURE AND WALL STRESS (PERCENT ОЕ 0.2% 
ТЕО ТЕО. Lux CUM I iS d gau cun nes 21 


TEST AREA POTENTIAL DROP NORMALIZED WITH RESPECT TO THE 
REFERENCE PARES FOR. EIECURE 5-2,2 ГТ ое. TZ 





Srl 
5212 
Sika]. 3 


223 
5-24 
5-23 


SAMPLE TEMPERATURE AND WALL STRESS (PERCENT OF 0.2% 


MIES TRES II A a e УЗ 
POTENTIAL DROP FOR TEST AND REFERENCE ÁREAS .................. 74 
NORMALIZED POTENTIAL DROP FOR БІСПЕН 5-5.................... Z4 


STEREOSCOPIC COMPOSITE OF SAMPLE SURFACE SHOWING CRACK 
EMANATING FROM BENEATH THE REFERENCE АВЕА РЕОВЕ ............... 793 


SEM COMPOSITE MICROGRAPH OF THE REFERENCE AREA FRACTURE 


SURFACE O O о си ево ван 75 
SEM COMPOSITE MICROGRAPH OF THE OPPOSITE SIDE OF THE 

KEBERENCE AREA, MPRACTURE. GUREACE . 0... ME: оао ое ооо ан 76 
SEM MICROGRAPH OF REGION 1 CLEARLY SHOWING IGSCC IN 

IN cis aia arta A Re ERR, EE e 76 
SEM MICROGRAPH OF AREA 2 SHOWING DUCTILE FAILURE REGION ....... du 
DUCTTER TE SCCETRANSTTITONSBRECLON DSG 9 0 О cn Z7 


SEM CoMPOSITE MICROGRAPH OF IGSCC FRACTURE SURFACE 
SHOWDINEIBERTETICALSCRACH  ERONT SE a a 000% 79 


NORMALIZED ACPD FOR SAMPLE SHOWING CRACK INITIATION 


PENIS BASED ONFENEREASING Ока POTENTIAL. DROP... 3.60 + ses ose ans 79 
SCHEMATIC OF TUBE SAMPLES WITH FATIGUE PRE-CRACKS IN 

ESTATE. ne ren are ee 80 
SAMPLE TEMPERATURE AND APPLIED WALL STRESS (PERCENT OF 

ООО ЕНЕС ee ee TS 81 
POTENTIAL DROP VERSUS TIME FOR TEST AND REFERENCE AÁREAS......... 82 


STEREO MICROSCOPE IMAGE OF THE THROUGH-WALL CRACK SEEN 
ARABE DAS ESOR  DAESDIACHAINED NOTCA соо о ооо ооо 83 


Low POWER STEREO MICROSCOPE IMAGE SHOWING THE RELATIVE 
POSITION OF THE THROUGH WALL. CRACK TO THE PROBES ....... +... u. 83 


NORMALIZED POTENTIAL DROP (TEST AREA WITH RESPECT TO THE 
REFERENCE AREA) SHOWING AN INCREASING SIGNAL FROM THE 


О Е A ME ee ee en 84 
SEM COMPOSITE MICROGRAPH OF FRACTURE SURFACE CLEARLY 

АРА ет 85 
MAGNIFIED SEM MICROGRAPH OF IGSCC AREA OUTLINING Two 

POSSIBLE СВАЛОК (GROWTH ARRAS еее 85 
MICROGRAPH OF THE PRE-IATIGUBEMNEEA l.l. 2433 nen nen 86 
MICROGRAPHUOF THE DUCTILE БАГОВ ВОО ое рае. 86 


MicROGRAPH SHOWING THE PRE-FATIGUE TO IGSCC TRANSITION 
РР ее 86 





А- 1-3 
A-1-4 
2-1 
29-2 


À-3-4 
AS => 
A3-6 
Ago / 
A-3-8 
A239 


A=3-10 
yl 
Dog 12 
A913 
A-4-1 
A-4-2 


CALCULATED TE OTENTIAL ПЕОР МЕБСОПЕ АСТПА, .................... 
CALCULATED P EOTENDIAL DROP VERSUS ACTUAL e oa een unse 


ALLOY 600 TUBE CRACK GROWTH VERSUS STRESS INTENSITY 
ЕО НАСОС О а еее 


COMPARISON OF CRACK GROWTH DATA VERSUS STRESS INTENSITY 


(CR) RR ARI IU EA. v he 
Du» USDODECRACKIOROWIH RATES VERSUS PH oe aa nee se ee 


HEATUP 950726 SHOWING OSCILLATIONS OF THE HEATER 
A A TT DEM 


HEATUP PLOT SHOWING CONTRAST IN TIME BETWEEN 
BERIRCHEATTONFAÄND STATIC MODE E 2. 1 Ао. 


ВОО ПАСЕ TO "O-RING FROM TESD 950903... Was о с се. а 
ОСОО СУ ЕИ р бо 


SAMPLE PRESSURIZATION CONTROL BOARD TUBING AND TEMPLATE 
DV WE Xt MEE naue NN aA e aes is 


RECIRCULATION AND STATIC SYSTEM CONTROL BOARD TUBING AND 
РЕА РЕ ЕТА она 


ENIECIMON@ Oe oTEM (COMPONENT = lc ое. порно ооо 


EMERGENCY DEPRESSURIZATION AND SAFETY COLLECTION SYSTEM 
COM WEN E ааа е 


ДЇ @®УЛ\УЕШН ЕКА» ЕТА aos erras OR Meme eee arte 
AUTOCLAVE HEAD AND SAMPLE ASSEMBLY DETAIL... ооо стон 
BESES AMPE E ASSRMBDEY CDETATES. «ctr eer eterne ntur re ve e 
IDETATILEDOGONTROE WIRING SCHEMATICT. or efc s ea I ERE ETUR ERIS e FORK woes 
NOTCH DEPTH VERSUS SURFACE CHORD LENGTH еее вена o лесе к 


SCHEMATIC OF FIXTURE USED WITH THE MTS TESTING SYSTEM TO 
INDUCE EATICUE CRACKS IN THE ТЕСТ SAMPLES ЕИ Ос. 


SIDE PHOTO OF SAMPLE AND FIXTURE MOUNTED IN THE MTS 
м о. 


FRONT VIEW OF SAMPLE AND FIXTURE INSTALLED IN THE MTS 
BRUM REMISE MEL cuu wow S CERT ET SD E ER UE RUN QS TA 





List of Tables 


Sl 
922 
DEI 
52 
AIl 


ЕА 
22-2 
Ae 3 
А-3-1 
41-2 
А22-3 
А-а 
Bao = 5 
A= 3-6 


Д-3-7 
B 


MECHANICAL PROPERTIES OF HEAT 96834 


CHEMICAL COMPOSITION OF ALLOY 600 


оо ө ө ө ө ө ө ө ө © © © © © © ә е ө ө ө ө ө ө ө 


TABULATED RESULTS FOR STRESS INTENSITY AND GROWTH RATES........ 
TABULATED RESULTS FOR STRESS INTENSITY AND CROWTH RATES........ 


CN9000A PID CONTROLLER PARAMETERS FOR STATIC AUTOCLAVE 
а 


STATIC OPERATION STARTUP VALVE LINEUP 


RECIRCULATION OPERATION STARTUP VALVE LINEUP 


INJECTION SYSTEM VALVE LINEUP 


ALLOY 600 TEST PLATFORM COMPONENTS SUMMARY 


SAMPLE PRESSURIZATION SYSTEM COMPONENT ID.. 


STATIC SYSTEM COMPONENT IDENTIFICATION 


RECIRCULATION SYSTEM COMPONENT ID 


INJECTION SYSTEM COMPONENT ID 


DEPRESSURIZATION AND SAFETY COLLECTION SYSTEM COMPONENT 
ID 


ооо о о о о о о о о о ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө ө o. ө е ө ө ө ө 


CONTROL SYSTEM WIRING DETAIL 


TEST DATA WITH ERROR CALCULATIONS 


10 





i1. Introduction 


Since 1967, most pressurized water reactor steam 
generators have been constructed with tubes made of alloy 
EUN Alloy 600 is a strong, corrosion resistant nickel 
based alloy. However, environmentally assisted cracking 
(EAC) has occurred within tube sheets, just above tube 
sheets and in tight radius U-bends in this material. 

Today, steam generator tube degradation is the life- 
limiting phenomenon in pressurized water nuclear reactors 
with U-tube steam generators. Comprising over half the 
primary to secondary pressure boundary surface area, the 
Steam generator tubes are of critical importance to reactor 
safety. Extensive investigations by numerous researchers 
have failed to produce a cohesive failure theory for the 
cracking phenomenon, known as Stress Corrosion Cracking’. 
Stress Corrosion Cracking has been observed on both the 
primary and secondary side of the steam generator tubes, 
елап [ат Stress Corrosion Cracking (IGSCC) on the 
secondary side periodically occurs at sites of Intergranular 
Attack (IGA), though either IGA or IGSCC can occur 
independently. 

Because IGSCC is of such concern for reactor safety, 
frequent shutdowns for steam generator tube inspections must 
occur to ensure tube integrity. Each maintenance period not 
only costs millions of dollars in lost power generation, but 
many man-rem of maintenance personnel exposure. Development 
of an on-line steam generator tube monitoring system would 


decrease the periodicity of shutdowns for tube inspections 


! J.T. Adrian Roberts,Structural Materials in Nuclear Power Systems, 


(New York: Plenum Press, 1981), р. 340. 
Herbert H. Uhlig and R. Winston Revie, Corrosion and Corrosion 
Control, 3rd Ed, (New York: John Wiley & Sons, 1985), p. 366. 


i 





and provide a more reliable warning of impending tube 
failure. 

The scope of this thesis is to provide a proof of 
peanciple for detection of stress corrosion cracks in actual 
steam generator tube material under high temperature, 
pressure and corrosive environmental conditions utilizing 
previously developed alternating current potential drop 
(ACPD) crack detection techniques. If successful, the 
techniques used to detect the stress corrosion cracks can be 
expanded and refined into an on-line system in the future. 

In order to demonstrate a proof of principle for crack 
detection by ACPD techniques in a high temperature, 
corrosive environment, an autoclave system with both static 
and recirculation capability was constructed. Under in-situ 
conditions, steam generator tubes require years for stress 
corrosion cracking to occur. To expedite the process, tube 
specimens were exposed to highly caustic conditions (10% 
sodium hydroxide and 0.1% Na,CO, added to subcooled 
deionized water) at 315°C. Boiling was prevented by 
maintaining pressure greater than saturation. Prevention of 
boiling is critical to preventing local concentration of 
Caustic . 

The specimens were internally pressurized to obtain a 
hoop stress up to 145% of yield stress. Based on previous 
Studies of stress corrosion on alloy 600. this allowed crack 


WR : : 3 : А 
initiation in weeks . Further reductions in crack 


? M. Payne and P. McIntyre, “ Influence of Grain Boundary Microstructure 


on Susceptibility of Alloy 600 to Intergranular Attack and Stress 
corrosion Cracking., Corrosion=NACE, XLIV.No. У. (Мау 1988), pp. 314- 
319. 


12 





initiation time were realized by polarizing the sample to 
+150 mv with respect to a nickel electrode’. 

Crack detection was achieved using ACPD techniques 
previously developed’. Pick-up probes were attached on 
either side of an expected crack initiation site. An 
additional set of pickup probes were attached away from the 
test area to serve as a reference signal. The site of crack 
initiation was controlled by plating the tube specimen with 
nickel in every area but the crack initiation site. Probe 
spacing, frequency and current were determined empirically 
and through prior analytical results to yield optimum AC 
potential drop sensitivity. 

Static system tests were performed to (i) test system 
imeprity to the caustic solution, (ii) determine the 
ability of the selected temperatures, pressures and solution 
pH to initiate specimen cracking in a reasonable time frame, 
(iii) verify the ability of the system to initiate stress 
corrosion cracking in the specimen and (iv) verify the 
ability of the ACPD system to identify crack initiation and 
growth under these conditions. Subsequent testing obtained 
stress corrosion cracking data for the tube specimens for 


comparison to data available in the literature. 


* J. B. Lumsden,, S.L. Jeanjaquet, J.P.N. Paine and A. McIlree, 


"Mechanism and Effectiveness of Inhibitors for SCC in a Caustic 
Environment", (Seventh International Symposium on Environmental 
Degradation of Materials in Nuclear Power Systems - Water Reactors, Vol 
Il EsrSckenrjdoe, СО: NACE International. August 7-10,1995), pp. 317-325. 
?^ R.G. Ballinger and I.S. Hwang, "Characterization of Microstructure and 
IGSCC of Alloy 600 Steam Generator Tubing", (Final Report, EPRI TR- 
101983, Palo Alto: Electric Power Research Institute, (February 1993),p. 
22865 


3 








2. Background 


2.1 Nuclear Steam Generator Materials and Environment 


The first nuclear steam generator tubes were constructed 
using austenetic stainless steel'. Corrosion related 
degradation occurred at a relatively rapid rate. The next 
generation of generators were built with alloy 600 tubing and 
an evolving water chemistry control program . Though the 
general corrosion problems were mitigated to a great extent, 
stress corrosion and other localized corrosions continued to 


(охехел SS 


2.1.1 Steam Generator Construction 


The tubes of a typical steam generator used in nuclear 
applications are of either a recirculating, or U-tube, design 
(Westinghouse, Combustion Engineering) or a once-through design 
(Once Through Steam Generator or OTSG manufactured by Babcock 
and Wilcox)”. A typical steam generator has approximately 
ШОС straight-through tubes (Babcock and Wilcox design) or 
approximately 8000 to 9000 U-tubes in recirculating designs to 
provide adequate surface area for heat transfer from the 
primary to secondary fluid. The tubes can extend in length to 
A meters (68 feet) in a straight through design. 
Approximately 15 support plates in both designs provide 
vertical and lateral support and vibration reduction over the 


tube length. The tubes are normally not welded onto the 


1 J.R. Cels, Caustic Stress Corrosion Cracking Studies at 288C(550F) Using 


the Straining Electrode Technique, Corrrosion-NACE, V34, No.6 (June 1978), 
198. 

James A. Adams and Eric S. Peterson, Steam Generator Secondary pH During 
A Steam Generator Tube Rupture, Nuclear Technology, Vol 102 (June 1993), p. 
205. 

У J.T.A. Roberts, Structural Materials in Nuclear Power Systems, (New York: 
Plenum Press, 1981), p. 323. 


14 





support plates and have clearances up to 0.25 mm (.010 in) 
between the support plate and tube to allow for thermal 
ОЕ Оо 7238 mm (1,5 in) thick tube sheet separates the 
primary from the secondary (one upper and one lower in a once- 
through design). The steam generator tubes are normally on the 
order of 12 mm to 18 mm in diameter and are heat fitted and 
tungsten inert gas welded into the tube sheet on the primary 
side. On the OTSGs manufactured by Babcock and Wilcox, the 
entire steam generator is stress-relieved in a furnace after 
final fabrication and welding of the hemispherical heads are 
completed”. 

As can be seen in Figure 
2-1, crevices between the 
secondary side of the tube 
sheet and tubes and on each 


side of the support plate 





tubes exist due to the en 


-———— ALLUY 600 TUBE 


Benserutetion deseribed. In ^ 


-CREVICE REGIDN 








addition to the crevice, these b m 
areas have high residual or 


working stresses due to the 





press fit or constraining 
Figure 2-1: Typical Areas of 


nature of the design. These Steam Generator Tube Failure By 

: | Stress Corrosion Attack 
crevices are typically the 
sites for intergranular attack (IGA) and stress corrosion 
cracking (IGSCC) and are the inherent flaw in steam generator 
design from a corrosion perspective. Additionally, residual 
stresses during U-tube bending can increase the localized 


stress levels many times higher than the hoop and radial 


stresses induced from the primary to secondary side 


P Babcock & Wilcox, Steam: its generation and use (New York: Babcock & 
Wilcox, 1975) р. 23-11. 


15 





differential pressure. These residual stress areas will be 


deseribed in more detail later. 


2.1.2 Steam Generator Chemistry 


The secondary side of nuclear steam generators operates in 
an alkaline environment with a pH typically in the 8.5 to 9.2 
range for recirculating steam generators and 8.8 to 9.6 for 
Babcock and Wilcox once-through steam generators. This pH 
range maintains a low general corrosion rate for all components 
in the steam generator since many components are either low 
alloy or carbon steel, including the steam generator shell. 

There are four principal additives used to maintain 
secondary chemistry’. The principal pH control additive is 
ammonia (NH,OH), selected for its compatibility with all 
materials used in steam generator construction. Ammonia is 
very volatile and only minimal residuals remain on the heat 
transfer surfaces and in crevices during power operation. This 
allows contaminants (such as chlorine or sodium) to concentrate 
in tube crevices and create a corrosive environment. Other 
additives, such as morpholine, boric acid and hydrazine are 
added to provide buffering in the crevices. Morpholine, being 
less volatile than ammonia, has provided protection of carbon 
steel components (steam and condensate piping outside the steam 
generator especially) from erosion/corrosion in areas of two 
phase flow. Boric acid has been added to reduce the occurrence 
of “denting” caused by the growth of magnetite in the tube 
support plate crevices in generators with carbon steel support 


plates and has also mitigated intergranular attack of alloy 600 


Е James P. Adams and Eric S. Peterson, "Steam Generator Secondary pH During 
A Steam Generator Tube Rupture," Nuclear Technology, CII, (June 1993), pp. 
304-306. 

ĉ Ibid, p. 305. 


16 





by high caustic/. Finally, hydrazine has been added as an 
oxygen scavenger. Although alloy 600 is not susceptible to 
BN ide stress Corrosion cracking, it is still susceptible to 
oxygen-associated localized corrosion. 

Chemistry in a steam generator crevice is not known. The 
case for a high caustic condition stems from much research and 
experimentation in this regime. R. Bandy, et.al., conducted 
tests on alloy 600 in high temperature caustic environments and 
concluded that the concentration of alkaline species was one of 
the more important factors governing intergranular corrosion . 
It should be noted, however, that although the general 
environment of the tests was alkaline, no measurements of the 
chemistry in the crevice area were performed and Bandy 
concluded that “a more thorough knowledge of the crevice 
chemistry is required”. S.M. Payne, et al., reported that 
impurities from the secondary water may concentrate in 
superheated hot leg crevices to 50% by weight of sodium 
hydroxide. Another possible source for free caustic 
Formation ie from ionic impurities introduced from corrosion 
products in the feed, condensate and chemical addition system 
and leakage in the condenser. These impurities include sulfur, 


phosphates, and chlorides, among others’. These impurities 


” “Boric Acid Application Guidelines for Intergranular Corrosion 
Inhibition,” EPRI NP 5558, Electric Power Research Institute (1984). 

James P. Adams and Eric S. Peterson, “Steam Generator Secondary pH During 
Pecteam Generator Tube Rupture,” Nuclear Technology, Gil, (June 1993), р. 
305. 

í R. Bandy, R. Roberge, and D. van Rooyen, “Intergranular Failures of alloy 
600 in High Temperature Caustic Environments," Corrosion - NACE", XVI no. 
IEEE MArchol985).pp. 142, 149-150. 

R. Bandy, R. Roberge, and D. van Rooyen, pp. 142-151. 

S. M. Payne and P. McIntyre, "Influence of Grain Boundary Microstructure 
on the Susceptibility of alloy 600 to Intergranular Attack and Stress 
Corrosion Cracking,” Corrosion-NACE, XLIV no. V, (Мау 1988), рр. 314-319. 

J.T.A. Roberts, Structural Materials in Nuclear Power Systems, (New York: 
Pienum Press, 1981).pp. 336-337: 


11 


17 








can concentrate in the low flow regions, such as the tube sheet 
crevice, or in sludge at the top of the tube sheet, and cause 
Keeallly high alkaline conditions in high heat flux, low flow 


regions (typically where intergranular corrosion occurs). 


2.1.3 Material Properties of Alloy 600 Steam Generator Tubes 


¡Aloy 600 is a nickel-chromium-iron alloy containing at 
Memote72, nickel (plus cobalt), 14 to 17% chromium, 6-10% iron, 
OS carbon (maximum, nominal carbon content is in the 0.08% 
range ) and small percentages of manganese, sulfur, silicon 
and copper. These concentrations are nominal and actual 
concentrations vary within this range. The importance of this 
variation and the small percentages of other materials added in 
the course of alloy 600 manufacture will play a role in the 
ability of a particular heat to resist intergranular attack and 
stress corrosion cracking. Typically in older vintage steam 
generators, following milling, the tubes are heated to a 
temperature to allow rearrangement of the individual atoms into 
a more stable structure. The tubes are then allowed to cool in 
the furnace. This process is known as mill annealing and 
softens the tubes to impart added ductility!*. Steam generator 
tubes conform to American Society for the Testing of Materials 
(ASTM) standard B-163 and American Society of Mechanical 
Engineers (ASME) Boiler and Pressure Vessel Code Section III 
(Nuclear Vessels) with a minimum tensile strength of 552 MPa 
(80000 psi) and minimum 0.2* yield strength of 241 MPa (35000 
psi) in a mill-annealed condition. Alloy 600 is ductile at 


room temperature with Charpy V-Notch impact strengths of 


i Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
John Wiley & Sons, 1980), p. 143. 

L. H. Van Vlack, Elements of Material Science and Engineering, 3rd Ed., 
(Reading, MA: Addison-Wesley Publishing Co., 1975), pp. 374-375. 


18 





approximately 250 Joules at typical steam generator secondary 
EndcEoberating temperatures of 224'C to 273'G (inlet and 

outlet, respectively)" for mill-annealed material. It does 
not embrittle after long exposure at high temperatures. 

Fatigue strength is affected by grain size and condition and by 
temperature, with typical values of fatigue strength greater 


than 325 MPa for 10° cycles” 


Nickel provides alloy 600 with a resistance to corrosion 
bvariety of organic and inorganic compounds. It is 
especially resistant to caustics and caustic solutions, the 
corrosion resistance being proportional to the nickel content 


in sodium hydroxide. 


Chromium is added to provide resistance to corrosion in 
high temperature oxidizing environments, and resistance to 
sulfur compounds (sulfur compounds act as catalyst poisons that 
ease access of hydrogen into a metal lattice resulting in 
cracking due to hydrogen embrittlement”). The effect of 
chromium additions to nickel is illustrated in Figure 2-2. The 
active region of Figure 2-2 is defined as the potential range 
at whieh significant corrosion occurs. The point at which the 
current suddenly drops to a value orders of magnitude lower is 
called the critical current density and the lower current 
density is the passive current density. The vertical region of 
Figure 2-2 where this lower current density exists is known as 


the passive region and the potential in this range is the 


> Neil E. Todreas and Mujid S. Kazimi, Nuclear Systems I (Bristol, PA: 
taylor € Francis, 1990), p. 5. 
: Inconel, Inco Alloys International (Huntington,WV: Inco Alloys 
International, Inc). 

Herbert H. Uhlig and R. Winston Revie, Corrosion and Corrosion Control 
(New York: John Wiley & Sons, 1985), p. 142. 


је 





passive potential’. Passivity for a metal is simply the state 
where the rate of dissolution in a given environment under 
steady state conditions becomes less as the electrode potential 
is increased than the rate at some lower potential”. At 
chromium concentrations greater than 9%, chromium in nickel 
significantly reduces the current density in the active 
erosion region. For chromium concentrations greater than 
10%, the critical current density is reduced, the passive 


potential widens and the passive current density is reduced”. 


ALLOY 600 
300*C, 20mV/min 


Reduced Cr decreases 
the passive potential 


10% Na OH 






= 0.8 

а 

» 

— d! 

a 96 Reduced Cr increases 

= the passive current density 
z 

ul 

5 0.4 

a. 


o 
ғә 


о Reduced Cr increases 
the critical current density 


“1072 io^! 10° 10! 10? 
CURRENT DENSITY (mA/cm?) 


Figure 2-2: Evans Diagram for Alloy 600 Showing the Effect of Chromium 
Concentration on Corrosion Current Density and Passivity. 


1? Herbert H. Uhlig and R. Winston Revie, Corrosion and Corrosion Control, 


ОТО Ed (New York: John Wiley & Sons, 1985), pp. 61-65. 
C. Wagner, Discussions at the First International Symposium on Passivity, 


Heiligenberg, West Germany, 1957, Corrosion Science, Vol 5, (1965). 9.751. 
e Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 


Schu Wiley 8 Sons, 1980), p. 136. 


20 





Chromium seems also to be the key factor in providing 
chloride stress corrosion resistance in alloy 600. This was 
demonstrated by Piron’, who added concentrations of NaCl to a 
H,SO, solution containing alloy 600 and nickel 200 which were 
made to passivate by controlling the potential until a 
breakdown of passivity occurred as indicated by a large change 
Bue passive current density. The critical NaCl 
concentrations were found to be an order of magnitude higher 
MI Udcs 600 than nickel 200 (2% vs. 0.1%, respectively) . 
Although the nominal concentration of 14% to 17% is specified, 
heat treatment of the material and differences during 
manufacture can cause localized regions at the grain boundary 
which contain concentrations significantly below the nominal, 
thereby altering the corrosion resistance of alloy 600%. The 
importance of this phenomenon in intergranular corrosion is 


further explained below. 


Iron is present in the stated percentage from the addition 
Sieterrochromium, instead of metallic chromium, during melting. 


Е Р 23 
Ferrochromium is used mainly to reduce cost”. 


Several precipitated phases can be present in the 
microstructure. These include titanium nitrides, titanium 
carbides, cyanonitrides, and chromium carbides, with the latter 
being the most prevalent because of the natural chromium 


content of alloy 600°". The particular concentration of 


р Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
пера Wiley & Sons, 1980), p. 139. 
i: Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
John Wiley & Sons, 1980), p. 143. 

Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
ШОйп Wiley & Sons, 1980), р. 142. 

Inco Alloys International, Inconel, (Huntington,WV: Inco Alloys 
international) p. 12. 


21 





carbides at the grain boundaries can affect the susceptibility 


of a particular heat to intergranular corrosion’. 


2.2 Stress Corrosion Cracking Theories 


The two prevalent theories which explain IGSCC involve 


einer anodic dissolution along the grain boundary or microvoid 


ev ЕО а Ее тате 


formation allowing crack propagation/^ 
implies, occurs along grain boundaries. Although transgranular 
cracking is possible, cases have been extremely rare and 
usually can be attributed to other factors influencing the 
crack propagation path. In post-mortems of tube failures from 
IGSCC, three common factors emerged. First, high stresses in 
the area of failure existed. Second, the material was 
susceptible to SCC. Finally, high temperatures influenced the 
damage rate^. In each theory, these factors influence a 


physical phenomenon which allows crack initiation, propagation 


Emdesventual failure. 


2.2.1 Film Rupture Theory for Crack Initiation 


Crack initiation is hypothesized to be the result of a 


film rupture mechanism which allows attack of the material at 


Ив. м. Payne and P. McIntyre, “Influence of Grain Boundary Microstructure 


on the Susceptibility of alloy 600 to Intergranular Attack and Stress 
Corrosion Cracking,” Corrosion-NACE, XLIV no. V, (May 1988), pp. 314-319. 


E^ R. Bandy, R. Roberge, and D. van Rooyen, "Intergranular Failures of Alloy 


600 in High Temperature Caustic Environments," Corrosion - NACE", XVI no. 
NE (March 1985), pp. 142-150. 


x Y, Shen and P.G. Shewmon, "IGSCC Crack Growth of Alloy 600 and X-750 in 


Steam," Corrosion, Vol. 47, No. 9 (September, 1991), pp. 712-718. 


e P.L. Andersen, ^Effects of Temperature on Crack Growth Rate in Sensitized 


Type 304 Stainless Steel and Alloy 600," Corrosion Science, Vol.49, No.9 
(September 1993), pp. 714-725. 


22 





Р 29 | А И 
асе атача boundary" . Alloy 600 in a caustic environment 
30,31 


passivates, primarily due to the high nickel content 
Passivation theory holds that a film, most likely of an oxide 
(Cr,0, is generally accepted in this case” although spinel 
oxides of Fe,0,, Cr40,, and Ni,0, have also been reported") or 
hydroxide composition, forms a thin protective layer on the 
surface of the material. The film lowers the current density 
significantly and therefore the corrosion rate is reduced 
significantly. The 14% to 17% chromium content, besides 
providing resistance to chloride stress corrosion, further 
lowers the corrosion rate. Experiments have shown that chromium 
concentrations in nickel greater than 9% significantly reduce 
the current density in the active corrosion region and for 
concentrations greater than 10%, the passive potential widens 


and the passive current density is reduced. 


Film rupture can be explained by two mechanisms. The first 


requires that a strain be applied. The second suggests a 


29 
R. Bandy, R. Roberge, and D. van Rooyen, “Intergranular Failures of Alloy 


600 in High Temperature Caustic Environments,” Corrosion - NACE”, XVI no. 
III, (March 1985),p. 149. 


Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
John Wiley & Sons, 1980), p. 157. 


is International Nickel Co., Inc., "Corrosion Resistance of Nickel and 
Nickel-Containing Alloys in Caustic Soda and other Alkalies," Corrosion 


Engineering Bulletin, CEB-2, 1971. 


E Wayne Z. Friend, Corrosion of Nickel and Nickel Based Alloys (New York: 
Jonn Wiley £ Sons, 1980). p. 165. 


s Warren E. Berry, Corrosion in Nuclear Applications (New York: John Wiley 
sons 1971) рр. 185-189. 


ur к. Sune, J. Koch, T. Angeliu, and 635. Маз. /Thesrfreet of’Grain 


Boundary Chemistry on Intergranular Stress Corrosion Cracking of Ni-Cr-Fe 
Alloys in 50 Pet NaOH at 140°C” Metallurgical Transactions A, Vol. 23A, No. 
10, (October 1992), pp. 2887-2904. 


н Маупе 2. Егіепа, Соггоѕіоп оЁ Nickel and Nickel Based Alloys (New York: 
John Wiley & Sons, 1980), p. 136. 


23 





breakdown of the film due to a change in the underlying 
material composition, thus altering the electrochemical 


Lu 36 
potential . 


Film rupture by a strain mechanism requires that a stress 
be applied to the material. The applied tensile stress levels 
can be as low as 10% of the yield stress’. However, larger 
stress levels are present in the failure areas when one 


examines residual stresses induced from manufacture processes. 


There are two basic steam generator tube designs used in 
pressurized water reactors today: U-bend (Westinghouse, 
Combustion Engineering) and straight-through (Babcock and 
Wilcox). The U-bend tube experiences the highest SCC rates 
primarily due to the stress relieving process which occurs in 
Babcock and Wilcox steam generators. Stresses present in U- 
bends are from applied and residual stresses (other stresses 
may be induced from corrosion or mechanical binding also). 
Typical operating stresses are on the order of 11 MPa in the 
Boop or circumferential direction and 5 MPa in the axial 


direction’. Both of these applied stresses are well below 172 


= Tea sung, .). Koch, Т. Angeliu and C.s. Was, “The Efiect of Grain 


Boundary Chemistry on Intergranular Stress Corrosion Cracking of Ni-Cr-Fe 
Alloys in 50 Pet NaOH at 140°C,” Metallurgical Transactions A, 
Vol. 235A ,No.10, (October, 1992). pp. 2895-2904. 


у Scaffer, Saxena, Antolovich, Sanders and Warner, The Science and Design 
б Engineering Materials, (Chicago: Irwin, 1995), p- 657. 


* Babcock and Wilcox, Steam/its generation and use, 39th Ed. (New York: 
Варсоск папа Wilcox, 1978). p. 31-22. 


E van Shah, D:B. Lowenstein, A.P.L. Turner. SSRo Ward. Т.А. Corman. Р.Е. 


MacDonald, G.H. Weidenhamer, "Assessment of primary water stress corrosion 
cracking of PWR steam generator tubes," Nuclear Engineering and Design, 
ош ал Nor 2-3, pp. 199-216. 


24 





to 345 MPa yield stress of alloy 600 following mill 


a 
annealing . 


-Flank 


Residual stresses, however, 
Extrados / / 


especially in areas of severe N E 
geometry changes (Figure 2-3), y / ) 


can be at or greater than the / / / 
| Intrades——4 / / A 
yield stress. Residual stresses ` | 


from the bending process tend to / / 


\ 


/ | /Irregular 


/ / 

/ / 

oecur directly adjacent to the / E 
L 
high curvature area. Tubes in fry ( 
> à 


a Sy Transition 
way, 

rows l or 2 have the tightest 

benderadius, on the order of 


Figure 2-3: Schematic of a 


steam generator tube 
in the extrados or flank can be showing areas of high 


deigh as 100 MPa or higher if residual stress 


55.4 mm. Residual hoop stresses 


ЕЕЕ сапе cold working was performed”. Of the approximately 
9000 U tubes in a typical steam generator, tubes in rows 1 or 2 


suffer the highest SCC failure rate, as expected. 


Santarini", Rios", and Shah/' all found that crack growth 


rate increased as the strain rate increased following an 


Р Inconel, Inco Alloys International (Huntington, WV:Inco Alloys 
International, Inc.). 


УД Velen onan Р.В Lowenstein, A.P.L, Turner, 5.К. Магға, ТА СсСогтап. P.E. 


MacDonald, G.H. Weidenhamer, “Assessment of primary water stress corrosion 
cracking of PWR steam generator tubes,” Nuclear Engineering and Design, 
ОЛИ №. 2-35. pp. 199-216. 


E R.B. Rebak and Z. Szlarska-Smialwska, "Influence of Stress Intensity and 


Loading Mode on Intergranular Stress Corrosion Cracking of Alloy 600 in 


Primary Water of Pressurized Water Reactors," Corrosion, Vol. 50, No. 5, 
(Мау 1994). рр. 328-393. 
43 


К. Rios, T. Magnin, D. Noel, O. DeBouvier, "Critical Analysis of Alloy 
600 Stress Corrosion Cracking Mechanisms in Primary Water,"  Metailurgical 
and Materials Transactions A, Vol. 26, No. 4, (1995), pp. 925-939. 


44 8. Shah, ор. сіс., рр. 199-216. 


25 





exponential relationship. This correlation supports a film 
rupture mechanism and is consistent with the observed stresses 


mama lire sites in U tubes". 


The second explanation for film breakdown is 
electrochemical. In 1984, Woodward ? identified a potential 
ПЕШ tot -750mV to -340mV(SCE) in which SCC initiated. 

Запа ,et al, theorize that chromium depletion (due to 
Пета л оп forming chromium carbide precipitates) or an excess 
of carbon migrating toward the grain boundary alters the 
electrochemical potential. This favors the formation of a less 
protective nickel-rich hydroxide film vice a protective nickel 
oxide. The nickel hydroxide film is stable only over a small 
tange of pH. If localized pH were acidic or highly caustic, the 
nickel hydroxide film would become soluble (unstable), leaving 


the underlying material unprotected. 


2.2.2 Crack Propagation by Anodic Dissolution 


Once a film rupture occurs, crack propagation in the 
matrix material can proceed. Bandy’, et al, suggested that 
film rupture is followed by anodic dissolution along the grain 


boundary and subsequent repassivation when the material is at 


E: VANES hah D.B. Lowenstein. ÀA.P.L. Turner, S.R. Ward, J.A. Gorman, Р.Е: 


MacDonald, G.H. Weidenhamer, “Assessment of primary water stress corrosion 
cracking of PWR steam generator tubes,” Nuclear Engineering and Design, 
9191324, №. 2-3, рр. 201-205. 


= J. Woodward, “Rapid Identification of Conditions Causing Intergranular 


Corrosion or Intergranular Stress Corrosion Cracking in Sensitized Alloy 
600,” Corrosion - NACE, XL, No. XII, (December 1984), pp. 640-643. 


P. IAK Súng. J. Koch., T. Angeliu, and 6.5. Was, “ihe Effect of Grain 


Boundary Chemistry on Intergranular Stress Corrosion Cracking of Ni-Cr-Fe 
Alloys in 50 Pct NaOH at 140'C," Metallurgical Transactions A, Vol. 23A, 
Noe lO. (October 1992), pp. 2887-2904. 


R. Bandy, R. Roberge, and D. van Rooyen, “Intergranular Failures of Alloy 


600 in High Temperature Caustic Environments,” Corrosion - NACE, XVI No. 
itive (March 1985), рр. 142-151. 


26 





the passive potential. Subsequent cycles propagate the crack 
along the grain boundary.  Bandy supported his conclusions by 
demonstrating that the rate of SCC is significantly influenced 
by the electrochemical potential. With samples in a potential 
range of +150 to 200 mV versus nickel (around the passive 

вешта! for alloy 600 at 300°C), grain boundary grooving 

occurred. However, for samples held at 0 mV versus nickel, no 


EOSCC occurred’. 


Chromium carbides are the most likely candidate for the 
cathode. The cathodic behavior of the carbides in caustic was 
demonstrated by Rebak, et al”. There are two major forms of 
carbide which precipitate at the boundary: M,C, and M,,C,. The 
major constituent of the M is chromium. The depletion of 
chromium at the grain boundary due to diffusion and subsequent 
precipitation as a carbide does not seem to influence the 
susceptibility of the material to ge Se 
carbon or other precipitates in the carbide as the major 
culprit. However, several researchers have linked carbon 
(оштете not only with crack initiation phenomenon, but crack 
growth rates. Johns and Beckitt found the susceptibility to 


IGSCC to be “critically dependent on the grain size and carbon 


: 55 
Content or the material. 


= R. Bandy, R. Roberge, and D. van Rooyen, p. 149. 


Ee Jang, "Effect of Sulphate and Chloride Ions on the Crevice Chemistry 


and Stress Corrosion Cracking of Alloy 600 in High Temperature Aqueous 
Solutions," Corrosion Science , Vol. 33, No. 1 (1992), pp. 25-38. 


к Rios, T, Magnin, D. Noel and O. deBouvier, "Critical Analysis of Alloy 


600 Stress Corrosion Cracking Mechanisms in Primary Water," Metallurgical 
and Materials Transactions A, Vol.26, No.4, (1995), pp. 928-929. 


"И D.R. Johns and F.R. Beckitt, "Factors Influencing the Thermal 


Stabilisation of Alloy 600 Tubing Against Intergranular Corrosion," 
Corrosion Science, XXX No. II/III, (1990), pp. 223-237. 


27 





The carbides that do form are precipitated out at the 
grain boundaries. Anodic dissolution would occur in the area of 
the grain immediately adjacent to the carbide. Once 
dissolution takes place, the material along the grain boundary 
and in the grain boundary is removed and a pit or crack is 
formed. With the cause of the initial film breakdown removed, 
repassivation can then take place. Like other mechanically 
induced cracks, stresses tend to concentrate at the crack tip. 
Seeain-induced crack propagation can then occur until a carbide 
particle is reached which can tend to pin the dislocations at 
the grain boundary and prevent further crack propagation. 
Again, electrochemical changes occur to break down the film, 


dissolution occurs and the process is repeated until failure. 


2.2.3 Crack Propagation by Microvoid Formation 


Pececond major theory for SCG crack propagation in alloy 
600 includes the formation of microvoids along the grain 
boundary. Generally, the theory holds that small voids from 
the formation of micro gas pockets, either hydrogen, or more 
likely methane, are formed in front of the advancing crack”. 
The reaction of hydrogen, formed by electrochemical processes 


at the crack tip: 
2H" +2e” >H, mc 
and carbon near the grain boundary, forms methane by: 


CAT TEN (9-9) 


53 Y. Shen and P. С. Shewmon, “ Intergranular Stress Corrosion Cracking of 


Alloy 600 and X-750 in High Temperature Deaerated Water/Steam," 
Metallurgical Transactions A, Vol. 22A, No. 8, (August 1991), pp. 1857- 
1864. 


28 





Lim and Raj” demonstrated that the nucleation of such 
microvoids “can be...aided by the impingement of slip lines 
along the grain boundary”, with a void spacing of 0.2um. It 
has also been observed that if carbides are present, the 
activity of the carbon at the grain boundary is reduced, the 
methane pressure is reduced and IGSCC is slower. If temperature 
increases, the stress, and by relation, the strain, increase. 
Stress-assisted diffusion, especially with stress concentration 
at the crack tip, may become an important factor in promoting 
diffusion toward the crack tip. Once hydrogen and carbon 
fmiiteract at the front of the propagating tip, small voids form 
which are eventually connected by the strain in the material, 
propagating the crack". As the crack propagates deeper into 
the material, voids continue to form and join ahead of the 


advancing crack, eventually leading to material failure. 


Although there are several other theories to explain 
stress corrosion cracking, the two theories presented here are 
ime only ones to incorporate observed microstructural, 
electrochemical and mechanical phenomena. Determining which is 
the actual mechanism is difficult. Each incorporates ideas 
which appeal to certain observed aspects better than the 
опеке. For example, dissolution theory does not seem to 
explain why primary water stress corrosion cracking is more 
prevalent than secondary side stress corrosion cracking, though 
microvoid formation seems to readily incorporate an 


explanation. In reality, there may be more than one mechanism 


F Y. Shen and P. G. Shewmon, " Intergranular Stress Corrosion Cracking of 


Alloy 600 and X-750 in High Temperature Deaerated Water/Steam," 
Metallurgical Transactions A, Vol. 22A, No. 8, (August 1991), pp. 1857- 
1864. 


?? Y. Shen and P. C. Shewmon, pp- 1857-1864: 


29 





Шеп астиа!!1у occurs to cause crack initiation and 


propagation. 


2.2.4 Impurities in Secondary Steam Generator Water 


The dissolution of the passive layer, and its importance 
to IGSCC described above, may be accelerated by the 
BEneroduetion of chlorides or other ionic impurities’ 

Chlorides are specifically addressed due to their ability to 
either prevent the formation of or to break down previously 
formed passive layers. In theory, chloride ions either increase 
the permeability of the formed oxide layer, or allow metal ions 
to pass more easily to the electrolyte if a chemisorbed layer 
dis present. The effect is the same in both cases, allowing 
corrosion rates to increase by reducing the potential below the 


passive potential. 


Seawater is often used to cool the condensate discharge 
from the turbines, which is fed back to a steam generator in a 
typical Rankine cycle. Any tube leaks in the condenser will 
ОШ chloride ions into the steam generator (seawater is 
approximately 19,500 ppm chlorides). Although a typical 
nuclear plant condensate system is monitored for chloride 
introduction, it is possible for levels below detectable to 
slowly concentrate chlorides in the steam generator. 
Typically, chlorides are controlled in the ppm range to avoid 
this concentration. However, this is a bulk fluid level and 


concentration of chloride ions in a crevice due to a low flow 


i Wayne Z. Friend,Corrosion of Nickel and Nickel Based Alloys (New York: 
John Wiley & Sons, 1980),pp. 136-142, 


je Herbert H. Uhlig and R. Winston Revie, Corrosion and Corrosion Control 


(New York: John Wiley & Sons, 1985), p. 74. 


30 





situation can result in levels in the hundreds or thousands of 


parts per million. 


sulfur appears to be a concern only in acidic, oxidizing 
environments. Today’s steam generator chemistry, combined 
with deoxygenating water systems, prevent this type of 
environment from occurring. Despite this, IGSCC still occurs, 


Eus doubt on the importance of sulfur in promoting IGSCC. 


Free hydrogen is present in relatively large quantities on 
the primary side of the steam generator tubes, though in very 
minute quantities on the secondary side. The surplus of 
hydrogen results in a highly reducing environment on the 
primary side. Hydrogen inventories are maintained by (1) 
стон of H, into the primary, (2) water decomposition 
under a neutron radiation flux and (3) decomposition of some 
primary pH additives. Hydrogen is effective in removing oxygen 
by radiolytic recombination in the presence of a gamma flux. 
Rios, et al, found, however, that hydrogen can only enter the 
material if a crack has been previously initiated. This 
suggests that film rupture must occur before hydrogen affects 
the material. This would imply, along with the previous 
diseussion of microvoid formation, that hydrogen concentration 
indeed plays an extremely important role in SCC. If the 
microvoid formation theory is correct, the material in a higher 


concentration of hydrogen would have increased diffusion into 


a. Jang, "Effect of Sulphate and Chloride Ions on the Crevice Chemistry 


and Stress Corrosion Cracking of Alloy 600 in High Temperature Aqueous 
Solutions," Corrosion Science , Vol. 33, No. 1 (1992), pp. 25-38. 


iun. Rios, T. Magnin, D. Noel, O. DeBouvier, "Critical Analysis of Alloy 


600 Stress Corrosion Cracking Mechanisms in Primary Water,"  Metailurgical 
and Materials Transactions A, Vol. 26, No. 4, (1995), pp. 925-939. 


31 





the material (since the gradient would be definitely into the 
material). This could promote faster void formation and 
subsequent crack propagation. It is interesting to note that 
PWSCC occurs the most frequently of all SCC°°. This would then 
support the relatively larger SCC failures on the primary side 


versus the secondary side. 


2.3 Role of Inhibitors 


It is speculated that the addition of various compounds to 
the secondary steam generator environment may mitigate or 
terminate intergranular stress corrosion cracking. Recently, 
several researchers have investigated the use of Titanium 
Emseeunds (TiO,, Tyzor, TiLAC, and Ti(Bu),), cerium salts, and 
Zine compounds’. Though results indicate that indeed crack 
mitigation or the prevention of the onset of SCC is possible, 
the exact mechanism of interaction is still unknown. 

The majority of the laboratory testing has been performed 
in caustic environments which simulate the pH of an operating 
steam generator (pH=10) at a temperature of approximately 
5152. Typical sample geometries include C-rings or standard 
ASTM tensile specimens machined from alloy 600 stock provided 
by the manufacturers of nuclear steam generator tubing. Recent 
testing, as well as earlier results”, demonstrated promise 


using titanium compounds over other possible additives. The 


= ан. D. B. Lowenstein. A.P.L. Turner. Б.В. Мата, ЛА Corman TELE, 


MacDonald, G.H. Weidenhamer, “Assessment of Primary Water Stress Corrosion 
Cracking of PWR Steam Generator Tubes,” Nuclear Engineering and Design, 
cs Ме 2-33 рр. 199-216. 

E. Bümsden. s.L. Jeanjagüuet; J.P-N. Paine and 2.2Meriree, “Mechanism 
and Effectiveness of Inhibitors for SCC in a Caustic Environment", Seventh 
International Symposium on Environmental Degradation of Materials in 
Nuclear Power Systems - Water Reactors, Vol. 1 (Breckenridge, CO: NACE 
lunteruational, August 7-10,1995). pp. 317-323, 

2 J.B. Lumsden and P.J. Stocker, "Inhibition of ICA in Nickel Based Alloys 
in Caustic Solutions", Corrosion/88, Paper No. (Houston, TX: National 
Association of Corrosion Engineers, 1988). 


32 





сетио ту of the particular additive appears important” 
however. Measurement of the potential and/or applied potential 
is also performed to characterize the electrochemical 
environment. 

Results indicate that the inhibitors may react by one of 
two methods detected thus far. In several cases, the 
inhibitors act to raise the potential above which SCC will 
occur. The IGSCC potential region appears to be +150 to +200mV 
versus nickel and occurs at the greatest rate at the passive 
potential”. Addition of titanium compounds increases the 
threshold potential by a minimum of 50 milli-volts and in a few 
cases prevented SCC occurrence within the test period”. In 
other cases, the inhibitor is thought to stabilize the passive 
film” preventing the onset of SCC by temporarily or 
permanently delaying the onset of film breakdown. Both 
hypotheses correlate well with the film rupture/anodic 


isso lutlion theory for SCC. 


ыт. Miglin, J.V. Monter, C.S. Wade, M.J. Psaila-Dombrowski, and A.R. 


McIlree, “SCC of Alloy 600 in Complex Caustic Environments”, Seventh 
International Symposium on Environmental Degradation of Materials in 
Nuclear Power Systems - Water Reactors, Vol. 1 (Breckenridge, CO: NACE 
Weermational, August 7-10,1995), рр. 277-290. 

R. Bandy, R. Roberge, and D. van Rooyen, “Intergranular Failures of Alloy 
600 in High Temperature Caustic Environments,” Corrosion - NACE, XVI No. 
Ber (March 1985), p. 149. 

4 J. B. Lumsden, S.L. Jeanjaquet, J.P.N. Paine and À. McIlree, "Mechanism 
and Effectiveness of Inhibitors for SCC in a Caustic Environment", Seventh 
International Symposium on Environmental Degradation of Materials in 
Nuclear Power Systems - Water Reactors, Vol. 1 (Breckenridge, CO: NACE 
International, August 7-10,1995), pp. 321-322. 

? M.T. Miglin, J.V. Monter, C.S. Wade, M.J. Psaila-Dombrowski, and A.R. 
McIlree, "SCC of Alloy 600 in Complex Caustic Environments", Seventh 
International Symposium on Environmental Degradation of Materials in 
Nuclear Power Systems - Water Reactors, Vol. 1 (Breckenridge, CO: NACE 
Шігегпасіспайі, August 7-10,1995), р. 281. 


23 





2.4 Crack Measurement by Potential Drop Methods 


recent years two primary methods of detecting “small” 
cracks have emerged. The first involves the use of a constant 
frequency, or DC current, the second an alternating frequency, 
БН current. In each case, the initiation of cracks in the 
material results in a measurable potential drop when probes are 


placed across the crack initiation site. 


2.4.1 Theory of DC Potential Drop 


Direct current potential drop was initially used for crack 
detection due to the relatively straightforward application of 
drcctocurrent theory to metallic conductors. In direct 
current flow, the current travels along the conductor from 
injection point to pickup. The resistance of the conductor is 
dependent on the material's resistivity, p, which is a function 
of temperature, the conductor length, /, and area, A. 

Applying Ohm's Law and the relationship for resistance: 


_1рЁ 
А 


Assuming for a moment that the resistivity is constant with 


E (2330 


temperature, the reduction of cross-sectional area due to 
crack penetration results in an increase in the potential 
(assuming the conductor is not expanding in the direction of 
£). As can be seen, this simple relationship allows the 
detection of crack initiation and growth based on a simple 
reduction of area. 

On closer examination, several difficulties arise which 
Make the application of DC potential drop systems inferior to 
the use of AC potential drop detection schemes. The resistance 
losses in the conductor during current transmission cause 


hesting according to: 


34 





Losses p^ Rp (2-4) 
Ás temperature increases for nickel alloys, the resistivity 
increases resulting in an increase in potential. Additionally, 
Ене Junction of the input probes with the conductor surface can 
result in electromagnetic fields. These fields, caused by 
thermal differences between the two materials (i.e. 
“thermocouple effect”), can be significant”. 

The two effects, the resistivity change due to temperature 
and the thermocouple effect, can be compensated for by 
employing two techniques. The first is to measure the 
potential across a similar conductor (i.e. same geometry, input 
probe and pickup probe spacing, etc.) which is placed at the 
same temperature as the conductor under test. The thermal 
effects of the potential change due to resistivity changes can 
be mathematically eliminated by normalizing the sample with 
respect to the reference. The second method, developed by 
General Electric” and employed in many DC potential crack 
measurement systems, reverses the applied current at a 
specified interval (e.g. every 0.5 seconds). This changes the 
polarity of the induced junction electromagnetic field. The 
measured potential drop initially is additive with the 
electromagnetic field during the initial half cycle and 
Embsequently subtractive on the next half cycle. The 
thermocouple effect is essentially eliminated by averaging the 


measured potential during each cycle. 


P. T. Baumeister, E. Avallone, T. Baumeister III, Mark's Standard Handbook 


for Mechanical Engineers, 8th Ed. (New York: McGraw-Hill Book Co., 

ШИЕ ор. 16-11-16-13. 

E Е. Catlin, D.C. Lord, T.A: Prater and L.F. Coffin, The Reversing DC 
Electrical Potential Method (Schenectady, NY: General Electric). 

Bu RF Catlin, D.C. Гога, Т.А. Prater and L.F. Coffin, тле Reversing DC 
Electrical Potential Method (Schenectady, NY: General Electric). 


35 





In addition to the thermal effects on potential 
measurement, the relatively low resistivity of alloy 600 in the 
temperature range of interest requires a relatively large 
current be applied to allow for a sufficient potential drop to 
occur between the pickup probes. From equation (2-3), the gain 
for a DC potential drop system is inversely proportional to the 
change in area resulting from crack growth: 

hn -A (2-5) 

V 4, 
Given the low resistivity of alloy 600 (inverse of conductivity 
from Table 1), equation 2-5 requires the application of 
currents normally in excess of 1 ampere for a sufficient gain 
in voltage from small changes due to crack propagation into the 
wall. 

Switching requirements, reference samples, and large 
applied currents are drawbacks to the application of using DC 
potential for an on-line monitor system. As will be seen, 
these drawbacks can be mitigated or eliminated and an increase 
in sensitivity to crack initiation realized with an AC 


potential drop system. 


2.4.2 Theory of AC Potential Drop 


Use of alternating current in a potential drop system 
allows a higher sensitivity to crack detection by taking 
advantage of the skin effect present in conductors. The skin 
effect distributes the current density to the outside of the 
conductor with little or no current flow on the interior. The 
depth of actual current flow is frequency and material 
dependent and must be calculated to optimize the sensitivity 
response of the system. A brief review of the applicable 


electromagnetic propagation theory is presented for background. 


36 








R.W.P. King developed the following derivation for current 
ШЕНЕУ distribution as a function of the radius in a tubular 
conductor’. Assuming a conductor of good conductivity 
surrounded on the inside and outside by “space” (i.e. vacuum), 
the potential functions from the Maxwell- Lorentz Equations for 


an electromagnetic field, 


ао (2-6) 
are applied. A is referred to as the vector potential and В 
is the phase constant.  Expressing equation (2-5) in 
cylindrical coordinates, the solution is made separable. The 


radial solution is the only solution of interest since the 
length of the overall conductor (i.e., wires plus tube) can be 
considered infinite with respect to the radius of the 
conductor. This permits the current density at any radial 
point, r, to be the same at any point, z, along the length. 


Differentiating the radial solution yields the Fourier 





Equation: 
QUO NE 
Lb тет ій ‚0 (2.72) 
ОХ ХОХ “ 


— 


where Е. jew. variable of integrationsender k ао 


— 
— 


conductor, k is equal to the phase constant, p, where p is 
defined as: 


B=(1-7)/Tf uo . (2-8) 


Solutions to the Fourier equation are well known Bessel 


Functions of zero order, first and second kind (J, and N,, 


respectively): 


А, (г) = Р, „(В үг) + D:N, (Br) (EOD 


? Ronold W. P. King, Fundamental Electromagnetic Theory, 2nd Ed. (New York: 
Dover Publications. 1963), pp. 350-356. 


ЭЎ 





A,(r) is evaluated for each region depicted in Figure 2-4. If 


(k,b)?<<1 (a good assumption for the diameter of tube and 
conditions in a vacuum), then Á (r) is equal to a constant. The 
Mindamental electric vector, E,. is related to the potential 


шестог by: 


= — jo k? 
E, (r)- 2 4 (r) (2-10) 
Cn _ 
and the current density is simply: а 
2 м 72 
Me TE (De) «А 
И b V A 
ай (+ VA 
Since A (r) is constant, m is equal D |, 


r 


Цео. Substituting (2-10) into (2- 





9), differentiating and evaluating the 2 
constants of integration at the Figure 2-4. Tubudar 


boundaries, r-b and r-c yields: Conductor 


E, () - Dj 8, 0N (B, b) - NB, AG, b) (2-12) 


The final constant of integration, D, is eliminated by 
taking the ratio of Е. (+) фо Е (с). Substituting the expression 
mor the current density given by equation (2-11) into this 
palo, integrating over the area of the conductor from b to c 


and rearranging yields the current density as a function of 


— 


Moral applied current, Г: 


по E (E) LON - ХОЛ | a, 
5 TEN? J (Bi DN, (B, 5) - NB, DB 5) 


J and N, are Bessel Functions of Еее оссе гаа аа 


second kind. Тһе skin depth is simply the inverse of the 


magnitude of В|: 


38 





1 
Jus 


Figure 2-5 plots the current density for various values of 


DE (2-14) 


Db through the tubular wall for the alloy 600 Heat 96834 sample 
tube dimensions. Since the material factors which influence 


D, cannot be altered except by substituting sample material, 


Normalized Current Density 





Normalized Distance from Outer Wall 


7 beta*b-15 
^* beta*b-20 
^"  beta*b-25 
7"  beta*b-50 


Figure 2-5: Current Density as a Function of Distance from the 
Exterior Wall of a Tubular Conductor for Various Values of pb 


Figure 2-5 clearly shows the dependence of the skin depth upon 
frequency. For highly ferro-magnetic materials, even low 
frequencies will generate an appreciable skin effect, while 
non-ferro magnetic materials (such as alloy 600) must use much 


higher frequencies to generate any skin effect. At values of 


D, b below 10, the current density approaches an even 


distribution, typical of direct current, as one would expect. 


39 





Cliven the current density as calculated from equation (2- 
12), the potential drop between any two points on the conductor 
71 
equals 


Ji 
E 2 I (R4 jo L) 2 —i, (2-15) 
б 


ШЕ D, b is large resulting in significant skin effects 


(generally greater than or equal to 10 as discussed above) and 
the wall is sufficiently thick compared to the skin depth”? (er 


b24d,), then it can be shown that the surface impedance, Z,, 


equals: 
21 cE 
PELO 216) 
| 1,(с) 
where z, is the internal impedance given by: 
D; 147 
p lE (ode 
2n co 42 
Кенет Еле, (2-16) into (2-15) and relating to (2-14) yields: 
R=0 1--- (per unit length) 218) 
б 


or, more simply, the impedance is proportional to e This 


allows the potential drop to be increased at the surface of an 
AC system by increasing the frequency of the signal for a 
constant current through the conductor, thus increasing the 
sensitivity to small impedance changes. 

Application of the above theory to a potential drop system 
illustrates the advantages of such a system. As a crack 


develops of depth a, the length of current travel along the 


E I. Verpoest, E. Aernoudt, A. Deruyttere, and M. Neyrinck, "An Improved 


A.C. Potential Drop Method For Detecting Surface Microcracks During Fatigue 
Tests of Unnotched Specimens," Fatigue of Engineering Materials and 
Desuetures, Vol. 3 (1981),рр. 206-209. 

Ronold W. P. King, Fundamental Electromagnetic Theory, 2nd Ed. (New York: 
Dover Publications, 1963), p. 357. 


40 





surface is increased by 2a. The potential drop increases by a 


2 
Ratio of (128) according to equation (2-14). The measured 


potential drop can be a significant change if the applied 
frequency is high resulting in high surface impedance (equation 
2-17). If the skin depth is large compared to the wall 
thickness however, the measured potential may be small owing to 
a relatively low surface impedance, though any skin effect will 
enhance the detection over resistance changes due to area 
changes required for DC system operation. The potential drop 
sensitivity to small changes in surface impedance of AC systems 
when applied to crack detection has achieved a sensitivity as 
small as 50um in crack depth”. 

There is a consequence of the application of high 
frequency AC signals which must be considered. It has been 


found”* during operation of AC potential systems that an error 


~ 


proportional to f vice the Jf is == 
induced on the measured potential. 
This voltage error is attributed to | | — = 


*inductive pickup". The higher the I 





frequency employed, the larger the 
inductive effect. The leads which 


Beaten toe the tubular conductor form 

Елашке 2 0: “Inductive 
Coil" Setup by Pick-Up 
generates an oscillating magnetic Wires on the Tube 


Беси (Figure 2-6). This coil 


field and associated voltage on the 


B. s. Hwang, "Embrittlement Mechanisms of Nickel-Base Alloys in Water" 
(Ph.D. dissertation, Department of Nuclear Engineering, Massachusetts 
Шс л шге of Technology. 1987). p. 101. 

F.D.W. Charlesworth and W.D. Dover, “Some Aspects of Crack Detection and 
Sizing using A.C. Field Measurements,” The Measurement of Crack Length and 
Shape During Fracture and Fatigue, ed. C. J. Beevers (West Midlands, 
WREEMAS, 1982), р. 258. 


41 





conductors (wire and tube). This inductive “load” can reduce 
the sensitivity of the system by raising the “noise” level of 
the received potential measurement. The higher potential level 
requires a greater change in 1, thus a larger crack 
penetration, for crack detection. For non-ferro-magnetic 
materials, this error can be near 100% of the applied 

voltage" 

In summary, the AC potential drop method has proved 
superior to the DC potential drop method for use in high 
temperature and pressure, corrosive environments. In DC 
potential drop, large currents are required to obtain a 
potential drop which is measurable. For sensitivity to crack 
size, the current must be stable to within 0.01%, which is 
difficult at the required high currents’. Additionally, 
thermoelectric effects from these currents result in a bias of 
the measured potential. 

Shortcomings of the DC potential drop technique are 
overcome by using an alternating current source. The AC 
current yields an increased potential drop for the same applied 
current, provides noise rejection and is more sensitive to a 
given crack size than DC potential drop. The ability to vary 
the frequency, limited only by available equipment, allows the 
sensitivity to be further improved by taking advantage of skin 
effect. 

Several shortcomings of the AC potential drop technique 
must also be considered. Use of the AC potential drop 


technique requires well-insulated probe leads, and sensitivity 


Z5 


Ibid., p. 258. 
76 P 


I.S. Hwang and R.G. Ballinger, "A multi-frequency AC potential drop 
technique for the detection of small cracks," Measurement Science 
Technology, Vol. 3,(1992),p. 63. 


42 





is a function of probe spacing.  Inductive effects, especially at 
higher frequencies, can result in induced voltages in the probe 
leads. These induced voltages can mask potential changes due to 
crack initiation and growth, thereby decreasing sensitivity. Thus 
a tradeoff must be made between induced voltages and increasing 
current densities for optimal crack sensitivity. Further, the 
non-magnetic properties of alloy 600 (as well as many other 
alloys) require the use of high frequencies to take advantage of 
the higher current densities present from skin effect. The 
larger skin effect depths require a reduction in the probe 
Spacing to maintain the same sensitivity as in a magnetic 
material. This obviously requires the area of crack initiation to 


be localized during any experimental observations. 


2.5 Present Research 


The literature clearly identifies many possible factors 
influencing ICSCC in alloy 600. Experiments conducted by the 
numerous researchers cited have presented IGSCC data for alloy 
600 based on standard ASTM test samples, C-rings or wires tested 
in a laboratory environment. Actual detection of IGSCC in-situ 
and the stress conditions present during initiation have not been 
measured. Additionally, no test data is available for steam 
generator tube samples subjected to laboratory experiments for 


comparison to these parameters. 


The focus of the present research was separated into two 
phases. During phase one, an experimental test apparatus capable 
of duplicating an in-situ steam generator environment and 
subjecting alloy 600 tube samples to various wall stresses was 
constructed. The goal of the testing was to prove crack 
detection in alloy 600 tubes under in-situ conditions was 


possible using an ACPD system. 


43 





In phase two, actual alloy 600 tubing crack growth rate data 
for various applied stress intensities was collected for 
comparison to the literature results. Use of the ACPD system 
allowed monitoring of the crack growth rate while subjecting the 


Бено various stress intensities. 


44 


4" 


arab әлат йзчоту 4шезе 0007 409 4 

| se? һал зо. Baw әш? 
қалға 212A nds: 3o sat am 

"из козады ‚эё 781 dsg A 





3. Experimental Apparatus 


An autoclave system provides the environment necessary to 
accelerate the formation of stress corrosion cracks on a sample 
tube from the normal in-situ steam generator environment. The 
system allows for monitoring of the sample tube with an 
Alternating Current Potential Drop (ACPD) system and control 
functions to maintain environmental conditions and safety. The 
system was designed to allow a sample tube to be internally 
pressurized to approximately 150% of yield strength and 
плете а: а temperature of 315£1'C. The operating fluid is 
0.1% Na,CO, and 10% NaOH by weight which equates to a pH of 10 
at 315°C. A similar system was developed to detect crack 
initiation successfully in a neutral environment at room 
temperature . Further, the system includes the capability to 
inject inhibitors and evaluate their ability to arrest crack 
growth. The autoclave system is illustrated schematically in 


Боге 3-1. 


3.1 Mechanical System Description 


The mechanical portion of the system consists of four 
components: (1) the autoclave and heating system (2) an 
autoclave recirculation loop (3) a static pressurization system 
and (4) a sample pressurization system. The autoclave system is 
designed to maintain the autoclave fluid subcooled in either a 
recirculation or static operating mode. The system may be 


operated in either mode with no mechanical alterations required, 


1 Т1 Soon Hwang, "Embrittlement Mechanisms of Nickel-Based Alloys in Water," 


PhD Dissertation (Massachusetts Institute of Technology, 1987),pp. 96-121. 


45 





aA 9piSjno OL 


E ps i Bm I = 
ых Е штән бицоо2 —- | 
il Ф ММ Мо Aiddng Buyoop —— | Ы г | 






ee 
ME 
Jj 
anc 
An 


Ге 


| | В LIN ТЕ 











Áiddns itv 
рәѕѕәзішод шоу 





een 


| 
De 


2 











1 ; | | J 
ШЕ ae р | штәң бицоо2 
2-озы 
|І 
Aiddng Buyoop 


= Е y HM 








eis 
E 
toas) (yov ] T 
А Bm T 
] 8 аы 


Autoclave System Showing Recirculation, Static and 


Figure 3-1: 


Pressurization Systems 


46 





Additionally, the mechanical system incorporates several safety 
and control features. Detailed construction drawings are 


contained in Appendix 3. 


3.1.1 Autoclave and Heating System 


A one gallon, nickel-200 
autoclave is used to contain 
the sample tube, heat the 
EInSude solution to 315'C and 
provide the required strength 
to maintain the caustic 
solution at a subcooled 
pressure. The vessel has a 
maximum allowed working 
pressure of 12.4 MPa (1800 
јео 315 C. The autoclave 
has a bolt-on head that has 
been modified with the addition 
of six one foot long nipples. 
The nipples are machined from 
nickel-200 bar stock that has 
been bored with a 6.4 mm (0.25 


in) hole. The nipples are 





meerteaq into the autoclave 


head and welded and have a 1/4 Figure 3-2: Autoclave with Heating 


қ | Mantle Showing Recirculation and 
National Pipe Thread (NPT) male ee 


thread machined into the 

opposite end. The nipples provide fittings for connecting the 
recirculation and static systems, penetrations for the ACPD 
electrical connections, sample pressurization tube and dual 
junction thermocouple. Each nipple has an external copper and 


bronze cooling jacket to maintain the stainless steel fittings 


47 





below 100°C, averting possible failure from caustic stress 
corrosion. The cooling jackets are provided with chilled water 


from the building chilled water system. 


The autoclave is heated by an external heating mantle. The 
mantle has a maximum heat output of 2700 watts from three 
elements in the mantle which are supplied with 250 volts each. 
Control of the mantle heat output is by turning any one or all 


che elements on and off. 


3.1.2 Recirculation Loop 


When operating in recirculation mode, the necessary liquid 
pressure and flow rate are developed by a Model 610 Pulsa-feeder 
pump. This pump is a positive displacement, wormgear reduction 
diaphragm pump capable of a maximum flow rate of 20 ml/min. 
ИБ) а а head up to 20 MPa (3000 psig). To simplify 
operation, the pump is made self-priming by locating the suction 


below a 200 liter storage tank. 


All tubing used to deliver the working fluid to and from 
ы Ил ос ауе is commercially procured 6.4 mm (0.25 in) x 1.6 mm 
(0.065 in) wall 316 stainless steel tubing rated at 70 MPa 
(10,000 psig) working pressure and conforming to ASTM standard 
A269/A213. 


All fittings used to assemble the tubing are Parker 
Biainiless steel CPI/A-Lok (working pressure rating ot 70 MPa) or 
National Pipe Thread (NPT) (minimum working pressure of 42 MPa 
(6100 psig)). 


All acolation valves are Parker ball valves: with 316 
stainless steel bodies and Teflon or Kel-F seats. The maximum 
working pressure of any isolation valve is 20 MPa (3000 psig), 


minimum. 


48 





A backpressure regulator, Tescom model 26-1724-24, on the 
fluid return line from the autoclave, maintains the desired 
fluid pressure in the autoclave. The regulator is stainless 
steel with teflon seats and is rated from 0 to 17.2 MPa (2500 


psig). 


A 200 liter (50 gallon) stainless steel tank with a quartz 
sight glass provides a supply and return reservoir. Purging and 
agitation of the fluid stored in the tank are accomplished via a 
E umm (0.25 in) nickel tube incorporated into the tank. An 
inert gas supply is attached to the end fitting and is bubbled 
through the fluid. The tank can be pressurized to a maximum 
pressure of 68 kPa (10 psig) and vented through a Parker CPI 
fitting at the tank top. The tank sits on a stand to allow easy 
access to the recirculation pump isolation valve and to a 


Conax® fitting, which allows internal potentials to be 


monitored. 


3.1.3 Static System 


In static mode operation, the system is pressurized with 
Grade 5 hydrogen. A gap of at least 25 mm (1 in) is maintained 
below the autoclave head as a gas space. The pressure source is 
a 40 MPa (6000 psig), 14000 liter (500 ft’) Grade 5 hydrogen 
bottle procured from a commercial vendor. A transfer hose, 
Normally used in hydraulic applications, which is non-conducting 
and rated to 70 MPa (10,000 psig) directs the gas to a 
regulator. Pressure in the tank is monitored by a mechanical 


gauge attached at the outlet. 


The pressure in the autoclave is maintained by a gas 
regulator obtained from Grove, Inc., model 15LHX. It has a 
stainless steel casing with stainless steel internal components 


and has a working pressure range from 0 to 41.3 MPa (6000 psig). 


49 





The regulator has a relief feature which allows pressure to be 
maintained within the desired range once set. Outlet gas from 
the regulator travels via a scrubber before reaching the 


autoclave. 


A scrubber is used to prevent hot caustic from reaching the 
regulator when gas is relieved from the autoclave to maintain 
the upper pressure setpoint. Hot caustic present in gas would 
quickly corrode the internal stainless steel regulator parts, 
resulting in regulator failure. The scrubber consists of a 
titanium 4 liter (1 gallon) autoclave (High Pressure Equipment 
Model BC-4) with a working pressure of 15.2 MPa (2200 psig). 
The autoclave is maintained cold and is filled 50% with 
deionized water. Returning gas passes through the water and is 
chilled by natural convection and conduction prior to entering 
the regulator for discharge to the atmosphere. During the 
cooling process, the caustic entrained in the hot hydrogen gas 


Heonwecondensed into the water. 


Two Parker stainless steel check valves with a 70 kPa lift 
pressure prevent short circuiting of the scrubber by the return 


gas. The check valves have a working pressure of 20 MPa. 


Tubing, fittings and valves used exclusively by the static 
system are identical to those described for use in the 


recirculation system. 


3.1.4 Sample Pressurization System 


Internal sample pressurization is accomplished with 
nitrogen. To obtain the required sample internal pressure, 
commercial Grade 4.8 nitrogen from a 8500 liter (300 ft’) gas 
bottle supplies a Haskel gas booster pump. The pump, Model AG 
152, is capable of reaching pressures of 140 MPa (20,000 psig). 


50 





The outlet pressure is controlled using compressed air and a 


regulator. 


All tubing in the sample pressurization system is 3.2 mm 
Q5) by 0.9 mm (0.035 in) wall 316 stainless steel and is 
rated to a maximum allowed working pressure of 100 MPa (15,000 
psig). A non-conductive hose, normally used in hydraulic 
applications, connects the sample with the pressurization 
system. The maximum allowed working pressure of the hose is 


10000 psig. 


All valves, except the diaphragm isolation valve, are 
Autoclave Engineer series 10V2, 316 stainless steel with allowed 
working pressures to 80 MPa (11500 psig). The diaphragm 
isolation valve is an air-to-close Autoclave Engineer Series 
10V2 with a maximum working pressure of 75 MPa (11000 psig). 
This isolation valve is normally open and is shut by 275 to 480 


kPa (40 to 70 psig) air pressure. 


Fittings are Autoclave Engineer low pressure fittings with 
a maximum allowed working pressure of 80 MPa, Parker NPT 
fittings with maximum allowed working pressures of 50 MPa (7100 
psig), and Boston Hydraulic fittings with maximum allowed 


working pressures of 70 MPa. 


3.2 Control Systems 


Autoclave and sample pressures are monitored by pressure 
transducers, Model P-605, manufactured by Omega Engineering. 
The internal autoclave temperature is monitored from a dual 
junction Type K nickel sheathed thermocouple installed through 
the autoclave head. The transducers and one junction of the 
thermocouple provide input to six model DP-41E meters, also 


manufactured by Omega Engineering, which provide visual readout 


51 





and control functions. Additionally, the pressure readings are 
recorded and displayed versus time on a Gateway 486 
microprocessor via a Hewlett Packard 3852A Data Acquisition and 


Control Unit. 


Heater control is provided by an Omega Engineering CN9000 
microprocessor controller. The controller is a proportional - 
integral-differential controller and is tuned to allow 
temperature to be maintained within a 1 degree Celsius band. 
Input for the controller is from the dual junction Type K 


nickel-sheathed thermocouple installed in the autoclave. 


3.3 Safety Systems 


The heater control and pressurization systems are 
interlocked to prevent startup if autoclave pressure conditions 
are not met and to shut down the system safely if autoclave 
pressure control is lost or a sample through-wall rupture 
occurs. The system logic electrically disconnects the autoclave 
heating system should subcooled conditions be lost, autoclave 
Penperature rise above 325 C, nitrogen pressure fall to а low 
specification (variable dependent on desired wall stress) or 
autoclave pressure rise above 13.8 MPa (2000 psig). 
Additionally, If the logie circuit initiates a trip, air is 
isolated to the Haskel booster pump and the sample nitrogen 
pressure dumped via the diaphragm valve, preventing over 
pressurization of the sample. A common collection system 
equipped with an air cooling coil and collection bottle prevents 
hot caustic from becoming airborne. A schematic diagram of the 
interlock system is shown in Figure 3-3. A detailed wiring 


diagram is in Appendix 3. 


The sample and autoclave pressurization systems are 


protected from over pressure by rupture disk assemblies. The 


D 





autoclave rupture disk is installed in the 6.4 mm tubing line 
between the recirculation pump and the autoclave and is not 
isolable from the static system when operated in that mode. It 
provides over pressure protection at 20 MPa. The Sample 


Pressurization system has a similar rupture disk which provides 


Protection at /0 MPa. 


+ 


115 VAC 
SERVICE POWER 


| 


505-1 


C-LP-1 
125VAC-S 


125VAC-LC 


NP-1-P6-NC 
а 
Jj EMERGENCY 
о DUMP 
ACP-2-P6-NC 
N 
Nm a а. 
Ф A о 
d ОЗЕ Ie е 
х 3 g 
2 SYSTEM OVERRIDES; Ф - S 
EN SWITCH о z 9 
ACP-1-P6-NC z M 
с? 
I 
о 
d ACT-1-P6-NC 7 
a 
z y 
E 
о 
АСТ-2-Р6-МС 
e 
N-S-1 о N-S-2 


Figure 3-3: Interlock Control System Schematic 


For personnel safety, the recirculation and pressurization 
systems are housed in a single cabinet open in the back and 
right sides for installation of services and maintenance. The 
cabinet also houses the Omega DP41 meters which are separated 
from the mechanical equipment and tubing by a Plexiglas spray 
shield. All wiring below the spray shield is contained in anti- 
corrosive, wet environment 13 mm flexible cable rated for this 


use. Polyethylene collecting pans are installed below the tank, 


53 





cabinet and pressure vessel to prevent caustic from reaching the 


ped floor. 


3.4 Test Sample 
3.4.1 Material 


The material selected for use in this work is a 2.2 cm diameter 
alloy 600 tubing, heat number 96834, fabricated by Babcock and 
Wilcox Company, with a 1.4 mm wall thickness. The tubing has 
been low temperature annealed at 927+14°С?. Important 
mechanical properties of this heat are contained in Table 3-1. 
This particular heat has demonstrated susceptibility to 


environmentally assisted cracking during in-service conditions. 


Table 3-2 illustrates the composition of the as-tested 
material compared to the chemical composition provided by the 
manufacturer and the limiting chemical composition provided by 
Inco Alloys International for Inconel Alloy 600 compositions. 
The chemical composition of the as-tested material was 
determined by electron dispersive x-ray spectroscopy (EDX) 


E youre 3-4). 


Rdiocure 3-5 illustrates the typical microstructure in three 
directions of alloy 600 Heat 96834, low temperature annealed 
material. The sample was prepared by polishing to a 1 jum finish 
and electro-etching in a 5% nitric acid- 95% methanol solution 
Ep volts for approximately 20 seconds. From Figure 3-5, this 


particular heat of alloy 600 shows an equiaxed grain structure 


ШЕ С. Ballinger and I.S. Hwang, "Characterization of Microstructure and 


IGSCC of Alloy 600 Steam Generator Tubing." Final Report, EPRI TR-101983, 
(Palo Alto: Electric Power Research Institute, February 1993), p. 2-1. 


54 





with an average grain size of 35um'. ЕЕ а SEM 
mierograph of the tubing oriented in the longitudinal (L) 
direction showing a large number of intragranular precipitates. 
These precipitates have been investigated" and found to be 
primarily carbides. Additionally, discontinuous carbides also 


reside at the grain boundaries. 


Table 3-1: Mechanical Properties of Heat 96834 










ee == 


Limiting 

Chemical ou Sco oO so 55 
Composition j max max max max | max 

B&W 96834 
Manufacturer : : Ds 0550726 О Ооо 
Composition^ 2 
EDX Analysis ; З 5 E 0.1 = © 5 5 
Composition 


* Composition not detected 





Table 3-2 : Chemical Composition of Alloy 600 


ШЕ С. Ballinger and I.S. Hwang, "Characterization of Microstructure and 


IGSCC of Alloy 600 Steam Generator Tubing",Final Report, EPRI TR-101983 
(Palo Alto: Electric Power Research Institute, February 1993), p. 3-56. 
Hr. s. Hwang, "Embrittlement Mechanisms of Nickel-Base Alloys in Water" 
(Ph.D. dissertation, Department of Nuclear Engineering, Massachusetts 
Institute of Technology, 1987), р. 187. 


^ Inconel, Inco Alloys International (Huntington, WV:Inco Alloys 
international, Inc.). 

E.G. Ballinger and I.S. Hwang, "Characterization of Microstructure and 
IGSCC of Alloy 600 Steam Generator Tubing", Final Report, EPRI TR-101983 
(Palo Alto: Electric Power Research Institute, February 1993), p.2-2. 


2/5 





EDX Composition for Alloy 600 (Heat B6LA) 


6000 


5000 . | | Ni 


- Run | 
#1 | 
|----. Run | 
#2 





Counts 
оо 
о 
о 
e 


1000 





0 100 200 300 400 500 600 700 800 900 1000 
Energy (keV) 


Figure 3-4: EDX of alloy 600 Heat 96834 (Low Temperature Anneal) 





0.10 mm 





Figure 3-5: Three Dimensional SEM Micrograph of Heat 96834 (longitudinal, L; 
transverse, T; short transverse, S). 


56 








Figure 3-6: Heat 96834 SEM Micrograph Showing Intragranular Precipitates 


3.4.2 Sample Construction 


The test sample is a 10 cm long section of steam generator 
tube with a cap welded to each end. A ceramic spacer is placed 
in the interior of the sample to minimize the pressurization 
volume. On one endcap, a 3.2 mm hole is drilled and a 3.2 mm by 
boum tube, approximately 50 cm long, is welded into the joint. 
This tube connects the nitrogen pressurization system to the 


test sample. 


Two sets of samples were manufactured. The first had a 2 
cm wide band around the center of the interior wall thinned to 
0.8 mm. This enabled higher wall stresses to be obtained at 
lower internal nitrogen pressures. The second set of samples 
had a fatigue crack induced at the test area of the sample. 


This pre-crack allowed accurate predictions of initial stress 


Э? 





concentrations to be made. A detailed discussion of this method 


can be found in the experimental procedures section. 


Figure 3-7 shows a simplified schematic of the two test 





samples. Appendix 3 contains the detailed schematics. 
m — — — End Caps 
a- o pog = 
| £ jor i enm =. 
Er ep. re 
| | 








ar cca == But 


— Sample Tube 


Sample with Inner Wall Thinned to 0.8 mm 

















— — ——— Internal paca Paz = 
— € „ус 
EE ty 
жоо ш сі 
' Е Ei И 
Sample Tube 


Sample Used for Fatigue Pre-Cracking 


Figure 3-7: Simplified Schematic of Test Samples Illustrating Machining 
Differences in the Tube Sections 


3.5 Alternating Current Potential Drop (ACPD) System 


The ACPD system is based on one developed by Hwang/'^. The 
actual system cabinet is shown in Figure 3-8. A schematic 
diagram is shown in Figure 3-9. Data collection is automated 
and no operator intervention is necessary once the system is 


started. One significant improvement is the use of a Gateway 


' Il Soon Hwang, "Embrittlement Mechanisms of Nickel-Based Alloys in Water," 


PhD Dissertation (Massachusetts Institute of Technology, 1987),pp.96-121. 
* R.G. Ballinger,and I.S. Hwang, "Characterization of Microstructure and 
IGSCC of Alloy 600 Steam Generator Tubing," Final Report, EPRI TR-101983, 
(Palo Alto: Electric Power Research Institute, February 1993),p. 2-54. 


58 





486 Personal Computer using LABVIEW®. This software provided a 


better graphics display and simplified final data analysis. 


The ACPD system provides a multifrequency capability 
through the use of a Hewlett Packard Multi-Frequency Function 
Generator (Model 3325A). The input signal to the test sample is 
peovided by a high stability AC current driver, Model 465-35, 
manufactured by Perry Amplifier, which combines an operator 
selectable DC input signal with the AC output of the Multi- 
Frequency Function Generator. The output signal from the test 
sample is amplified by a Perry Amplifier Model 675D Low 
Impedance Pre-Amplifier. The signal is discriminated against 
background noise by a microprocessor controlled, high 
sensitivity Princeton Applied Research Model 5301 Lock-in 
Amplifier and passed to the Gateway 486 PC through a Hewlett 
Packard Model 3488A Switch Control Unit. 


=E 
КІШІ 


ра; ES 
^ E 


[IB a": 
ШЇ 





Figure 3-8:  ACPD Test Stand 


59 





| НР 34884 в _— 
*SWITCH CONTROL DNIT 





















| | — . AC | | 
E | PERRY  @ DCPW 

| PRINCETON 5301, CURRENT PRE AMPLIFIER | 

| LOCK-IN AMPLIFIER DRIVES e 

| Ф e— ee Ф:- 
| 

| | | | T 
| HP 3325A | 55 Е 
- FUNCTION GENERATOR | |, 

| em | 
S 
486 PC | e | 
рум Форин 
ЅАМРІЕ 
OSCOPÉ z 
р e m 


Figure 3-9: ACPD System Schematic 


60 





4. Experimental Procedure 


The experimental program was divided into two distinct 
phases. During phase one, the ACPD system proof of principle 
was to be accomplished. In phase two, the focus was on crack 
initiation and growth on tubing. To expedite the SCC 
initiation and growth process, a fatigue pre-crack was induced 
in the samples for phase two. The system operation remains 
identical in both cases. Detailed step by step operating 


procedures are in Appendix 2. 


4.1 ACPD Proof of Principle Testing 


During phase one, the goal was to demonstrate the ability 
of the ACPD system to detect cracks in conditions similar to 
in-situ steam generator conditions. The system was operated in 
static mode for this phase of testing. Figure 4-1 shows a 
welded test sample. Approximately 2 cm of the interior wall 
was thinned to 0.8 mm to allow a greater stress range to be 
applied without exceeding the pressurization system 
specifications. Detailed construction drawings are in Appendix 


2. 


Test Area 





Figure 4-1: Welded Test Sample. 


61 





After welding, the sample was electroless nickel plated in 
all areas except a small area in which crack initiation was 
expected. The plating prevented SCC initiation at sites beyond 
that monitored by the ACPD test probes. Crack initiation and 
growth were expected in the axial direction since the hoop 
Stress is greater by a factor of two than the axial stress in 
the cylinder. The test site consisted of an area no greater 
than 1 mm across in the circumferential direction to improve 
crack detection sensitivity, by 3 mm wide in the axial 
direction, roughly the weld contact area of the output probes 
to the ACPD system. This area was masked off with Miccro Super 
XP 2000 micro-stop and allowed to dry 24 hours to ensure mask 


cohesion. 


Electrical isolation of the sample and ACPD probes was 
necessary to maximize output gain from the ACPD. Figure 4-2 
shows schematically the sample tube installation into the 
autoclave head. If grounds exist, current across the test site 
is reduced and overall potential drop is lowered. The sample 
was electrically isolated along the pressurization tube by 
encasing the tube in High Operating Temperature Teflon (TFE) 
(Alpha Wire Corporation, No. FIT-500). This teflon insulation 
was also used on all input and output ACPD leads to the sample. 
The sample was passed through one of the autoclave head 
nipples via an electrically isolable Conax® fitting (Conax 
Corporation catalog number EGT-125-A). ACPD system input (one 
set) and output (test and reference area) probe wires were 
similarly passed through separate nipples to prevent inductive 


interference and through Conax® fittings TG-24-A4. 


62 





- ACPD Pick-Up Wires Installed 
Through CONAX Fitting 
(input wires installed through another 


Pressurization Tube Installed CONAX fitting, not shown) 


Through CONAX Fitting 
(see Appendix 3 for details) 


Cut Away View of ~~~ 


Autoclave Head Nipple Autoclave Head 














—— Autoclave Head 








Zirconia Spacer (40 mm long) — < ACPD I Wi 
OP NEED nput Wires 


(sheathed with high temperature teflon) 





е | Ш E ACPD Pick-Up Wires 
(sheathed with high temperature teflo 


Nickel Plated Tube Sample 


Figure 4-2: Schematic of Sample Tube Installed Into Autoclave Head 


The input and output probe wires were then welded to the 


sample. The input probes were welded along the same axial 


plane as the test area at approximately +90 degrees azimuth. 


The reference area was chosen along the same axial plane 


between one input probe and the test area. Input wire spacing 


гіс спе reference area should be the Same as the test area to 


63 





Peovide initially identical potential drop signals. Figure 4-3 
shows the completed head and sample assembly. The entire 
sample, less the masked off test area, was then nickel plated 
using a plating procedure developed by Morra for surface 


specimen preservation (detailed procedure is in Appendix 2). 





Figure 4-3: Sample Assembled into Autoclave Head Showing ACPD Input and 
Output Wires, Thermocouple, Level Sensing and Nickel Reference Electrode 


Following completion of the plating process, the masking 
was removed from the test area. A small sharp surface scratch 
was placed in the test area to aid in stress concentration. 
The thermocouple (not electrically isolated), nickel reference 


electrode and several wires for fluid level measurement by 


*M. M. Morra, J. М. Morra and R. R. Biederman, “A Technique for the 


Preparation of Powders for Examination by Transmission Electron 
Microscopy”, Materials Science and Engineering, A124 (1990) pp. 55-64. 


64 





potential were installed in the head following the procedures 


ЕС Гог the ACPD wire installation. 


The caustic solution was prepared using 10% NaOH and 0.1% 
Ма,СО, by weight in deionized water. Тһе Ма,СО; was added to 
accelerate the corrosion process’. 2600 ml of solution was 
prepared and added to the autoclave to allow for a minimum of 3 


cm (1.5 inches) of gas volume at the top. 


The head assembly was installed and torqued to 170 to 190 
Teme 125 to 140 ft-lbs). Cooling tubes were attached to the 
head nipples. All external tubing connections for the 
static/recirculation system were attached to provide the 
hydrogen pressurization path. Electrical connections to the 
thermocouple and polarization system were completed. ACPD 
constant current transformer outputs were connected to the 
probe input leads. The reference and test area leads were each 
connected to a separate Perry preamplifier. The ACPD system 
was started and output signals observed for stability (noise, 
transients and signal level) using an oscilloscope and 


monitoring the computer output. 


The system and autoclave volume were purged of oxygen with 
grade 5 hydrogen. Purging consisted of flowing hydrogen at 15 
kPa (about 2 psig) through the Parker tee on the outlet tube 
and exhausting through each nipple vent (ACV-1 through 6) for 
approximately 15 minutes each. An alternate purge consisted of 
using the hydrogen supplied to the scrubber and again 


exhausting through each nipple vent for 15 minutes. The system 


? R. Bandy, R. Roberge, and D. van Rooyen, “Intergranular Failures in Alloy 


600 in High Temperature Caustic Environments,” Corrosion-NACE, Vol. 41 No.3 
March 1985),pp. 142-151. 


65 





passthen pressurized to 3.5 MPa (500 psig) through the scrubber 
by opening REG-2 and leak checked at all fittings. 


Detailed system startup procedures are in Appendix 2. The 


system hydrogen overpressure was raised to 12.4+0.5 MPa 
(1800£50 psig). The heating circuit was energized and 
mitoebave temperature raised to 315+1°C. Once at temperature, 


the ACPD signals were again monitored for stability. The 
desired stress across the wall was obtained by operating the 
sample nitrogen pressurization system to obtain the desired 
differential pressure. During this phase of testing, internal 
Mogen pressure was raised to approximately 45.5 MPa (6600 


psig) to provide a nominal wall stress of 140*. 


To aid in SCC initiation, the sample potential was 
polarized anodically with respect to the nickel autoclave. As 
seen in Figure 4-4, the corrosion current is significant at 
potentials greater than +150 mV versus nickel. In fact, 
previous testing has shown that SCC does not occur at 
potentials below +100 mV versus nickel’. Due to the relatively 
large surface area of the test sample, a DC constant voltage 
power supply (Hewlett Packard Model 6023A) was used. The 
positive terminal was attached to the sample pressurization 
tube and the negative terminal to the autoclave (counter- 
electrode). The potential versus nickel was monitored using 
the electrically isolated nickel electrode installed throush 


the autoclave head and a high impedance voltmeter (Keithly 


? 7. B. Lumsden, S. L. Jeanjaquet, J. P. N. Paine and A. McIlree, 


"Mechanism and Effectiveness of Inhibitors for SCC in a Caustic 
Environment," Seventh International Symposium on Environmental Degradation 
of Materials in Nuclear Power Systems - Water Reactors, Vol. 1, 
(Breckenridge, CO: NACE International, August 7-10,1995), pp. 317-325. 


66 





Programmable Electrometer, Model 617).  Potentials were 
maintained between +150 mV and 220 mV versus nickel during this 
phase of testing. If drift in the applied potential is noted 
during the course of the test due to the lack of a feedback 
loop to the power supply, adjustment to the applied potential 


must be made to prevent drifting outside the desired band. 








Potential (V vs Ni) 


- о == гаси о << ttt tt 
E-02 1.00E-01 1.00E*00 1.00E+01 | 








Figure 4-4: Polarization Curve for Alloy 600 at 300C in 10% NaOH* 


During phase one, the potential drop for the test area was 


monitored versus the reference area. The test was terminated 


“В. Scarberry, "Corrosion of Nickel Based Alloys", Conference Proceedings, 


American Society of Metals, Metals Park, 1985. 


67 





when a 50% rise in test area potential drop was indicated with 


no associated increase in the reference area potential drop. 


4.2 Crack Initiation and Growth Measurements 


Following successful completion of the ACPD proof of 
principle testing, phase two tests were conducted to obtain 
crack growth data from the steam generator tube samples. The 
ACPD system was used to detect crack initiation and monitor 
growth rate. Stress intensity calculations were made based on 
a finite element analysis of the crack geometry. The data 
obtained was then compared to data obtained from other 


researchers using non-tube samples. 


Crack initiation times were reduced by inducing axial 
fatigue cracks in the tube test area prior to inserting into 
the autoclave. Axial fatigue cracks were initiated at the test 
site by applying a circumferential load along one axial point 
Mette an MTS fatigue frame. The initiation site occurs at 90 
degrees to the load axis by machining a pre-crack with a small 
grinding tool. This increases the stress at this point beyond 
that experienced along the load axis. The site was monitored 
with a strobe and fatigue crack length measured. The fatigue 
growth was stopped when cracks were visible on each side of the 
pre-flaw. Measurements on several tubes showed that this 
produced an approximate elliptical crack front. The stress 
intensity factor can be calculated using a finite element 


analysis of Newman and Raju knowing the surface length of the 


? J. C. Newman Jr., and I. S. Raju, "An Empirical Stress Intensity Factor 


Equation for the Surface Crack," Engineering Fracture Mechanics, Vol. 15, 
EE UI-2(Great Britain: 1981), рр. 185-192. 


68 





a. | 


fatigue crack. Final stress intensity was less than 25 MPaYm. 


The fatigue pre-cracks were initiated on the sample tubes prior 


to welding the caps. 


Preparation of the sample, installation and operation of 
the system were identical to phase one. Pickup probes were 
carefully welded to each side of the pre-crack. The sample was 
plated ensuring adequate masking of the pre-crack area to avoid 
plating into the pre-crack. The caustic solution was identical 
to phase one and the sample was again polarized to +150 to 220 


mV versus nickel. 


At the conclusion of the test, the tube sample was removed 
and the axial ring encompassing the test area cut from the 
tube. The ring was fractured ductilely in tension using the 
MTS Test System to expose the IGSCC growth area. Measurement 
of the fatigue, IGSCC and ductile areas allowed calculation of 
the stress intensities for the applied stress. Growth rates 


were measured from the ACPD data. 


69 





5. Results 


SI ACPD Proof of Principle Testing 


The goal of this phase of testing was to prove the 
operability of the ACPD detection system to detect IGSCC 
initiation and growth under in-situ steam generator secondary 


conditions. 


The initial data showed the potential drop beginning to 
басе within the first 10 hours of the start of the test (Figure 
5-1). A steady increase in the potential drop continued for 68 
hours at which time a sharp increase in the signal occurred. 
The sample wall stress was reduced by decreasing the internal 
sample pressure to check the validity of the potential drop 
signal (Figure 5-2). A drop in the potential indicated the 
possibility that crack initiation had occurred. The sample 
remained in the depressurized condition for approximately 12 
hours. Repressurization of the sample demonstrated a return to 
the potential drop level previously recorded. The test was 


berminated at 140 hours when no further crack growth was noted. 


A close visual examination under the stereoscopic 


microscope revealed no cracking of the wall in the test area of 


the sample. However, circumferential cracking of one of the 
test area pickup probes was noted. The crack appeared to be 50% 
through the 0.5 mm diameter wire. Normalizing the potential 


drop data clearly showed a rise in the test area signal with 


Beepect to the reference area (Figure 5-3). 


70 





0.0095 


0.0090 


0.0085 


0.0080 
0.0075 é 
Test Area n 
i te 
cue n y 
0.0070 | 


0.0065 Reference 


0.0060 


Potential Drop (milli-volts) 


0.0055 


0.0050 


0.0045 


0.0040 





0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 


Time (hours) 


Figure 5-1: Potential Drop for the Test and Reference Areas. 


325 110 
100 
90 
320 = 
£ 
h 
- 
80 
щл 
с 
~ ч 
Ur т 
o 70 % 
0 315 = 
E Ae 
2 60 о 
м th 
u о 
5 M 
£ 50 У 
v 310 = 
G 4а С 
о 
Ui 
un 
я 
зо н 
fh 
Uu 
и 
305 d 
20 
Sample Temp 
© Vall Stress 10 
300 0 





О 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 


Time (hours) 


Figure 5-2: Sample Temperature and Wall Stress (Percent of 0.2% Yield 
Stress). 


Ze 





1.45 
1.40 
1.35 
1.30 
1.25 
1.20 
1.15 
1.10 
1.05 


90 
85 
80 
75 
70 
65 


60 


Normalized Potential Drop (Test Area/Reference) 


55 
50 


о о о O O0 ORO O O O 


45 


o 


40 
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 


Time (hours) 


Figure 5-3: Test Area Potential Drop Normalized With Respect To the 
Reference Área for Figure 5-2. 
Although the cracking did not occur in the sample, the test 
did show the ability of the ACPD system to detect crack 


initiation and growth in the caustic environment at temperature. 


Additional data to confirm ACPD operation under in-situ 


steam generator conditions is presented in Figures 5-4 through 


5-6. For this sample, the initial wall stress was set to 90% of 
yield stress. The sample was polarized to +225+5 mV versus 
nickel based on previous studies by Bandy, ЕЕ а Subsequent 


increases in wall stress in an attempt to expedite crack 


ER tiation occurred at the 400, 510, 600 and 740 hour points 


ER. Bandy, R. Roberge and D. van Rooyen, “Intergranular Failures in Alloy 


600 in High Temperature Caustic Environments," Corrosion-NACE, Vol. 41, No. 
Be (March 1985),pp. 144-145. 


72 





Шіспге 5-4). 


TInerementalr inereases were used te ensure 


pressurization system integrity and to attempt to find the 


lowest optimum stress for crack initiation. 


325 


320 


315 


310 


Sample Temperature (C) 


305 


300 





150 


100 


125 


50 


25 


Sample Temperature E 


O Sample Vall Stress 


(ss2128 PISTA %770 20 %) ввә115 TION ardues 


-50 


O 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 


Time 


Figure 5-4: Sample Temperature and Wall Stress (Percent of 0.2% Yield 


Stress). 


Two frequencies, 1 kHz and 10 kHz, were used at 800 mA and 


monitored by the ACPD system. The use of dual frequencies in 


these ranges allowed a relatively low frequency signal (which 


can be considered direct current (DC) from a skin effect 


viewpoint) and an alternating current (AC) signal, with some 


minor skin effect and minimal induction effects, to be used for 


detection and comparison. 


Figure 5-5 plots the results of the potential drop 


measurements. 


Although no significant increase in the test area 


Potential drop occurred, the reference area displays a 


73 





significant increase in both the 1 and 10 kHz signals. 


A sharp 


rise in the 1 kHz signal is easily seen at the 1400 hour point 


tae the test. 


The potential drop measurements were normalized 


for the reference signal with respect to the test area signal 


сиве 5-6). 


520 until test termination is clearly seen. 


Normalized Potential Drop {1 kHz) 


Potential Drop (milli-volts) 


Figure 5-5: 


10 





10 kHz Test Area 


ы sor fame O 
ее = -r as au RTV 
= = EC “ws, bap ante mM ee > z 








= = 


paa 


10 kHz Reference 


n SR 


1 kHz Test Area 
= = Par n E cc У 


Y 
R 1 kHz Reference pi» 


IP dd e e 





O 100 200 300 400 500 600 700 800 900 100011001200130014001500 1600 


Time (days) 


10 kHz Signal 


yw 
жоба” 


1 kHz Signal 





300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 


Time (hours) 


Figure 5-6: Normalized Potential Drop for Figure 5-5. 


74 
















Potential Drop for Test and Reference Areas. 


The steady rise in the 10 kHz signal from hour 


(2HX OI) doaq Tetqausqog paztTewz0N 





After operation for 1540 hours, the sample was 
depressurized and the autoclave cooled for sample inspection. 
Close visual examination under the stereoscopic microscope 
showed an approximately 2 mm long crack emanating from beneath 
the right reference area pick-up probe (Figure 5-7). Several 
smaller surface cracks were also noted. The tube was sectioned 
and the crack surface opened under tension. The area of 
intergranular stress corrosion cracking is well defined under 


DEN examination (Figures 5-8 and 5-9). 


р a и * 
Eminating From 
2 Spot Weld 


5 :% 





Figure 5-7: Stereoscopic Composite of Sample Surface Showing Crack Emanating 
From Beneath the Reference Area Probe. 


47% 


IGSCC | Du Spot Weld 





Figure 5-8: SEM Composite Micrograph of the Reference Area Fracture Surface. 


LD 





“Secondary TÉSEC Sites 
on Outer Surface 





Figure 5-9: SEM Composite Micrograph of the Opposite Side of the Reference 
Area Fracture Surface. Note the Secondary IGSCC Cracks Apparent on the 
Surface Near the Primary Site. 


Figures 5-10 and 5-11 are magnified micrographs of the 
areas indicated in Figure 5-8. The dimple effect of the ductile 
failure region in contrast to the easily distinguished grains 
and lack of deformation in the indicated IGSCC region are 
clearly seen. The area of interest is clearly the result of 
IGSCC, not ductile fracture. The boundary area micrograph in 


Биеке 5-12 І1ІТйеттагев the transition region. 


1SKU 1.50KX 6.67" 8600 





Figure 5-10: SEM Micrograph of Region 1 Clearly Showing IGSCC In This 
Region. 


76 





Figure 5-11: 





DN a: p 
ДУЗ. 





pone 


AL Mn 


15KU 1.50KX 5. aT 0601 


SEM Micrograph of Area 2 Showing Ductile Failure Region. 


"~. 54 


¥ 


+ 
~ 


15KU 300X 


Figure 5-12: Ductile-IGSCC Transition Region. 


77 





The results of the potential drop measurements and 
subsequent confirmation of IGSCC demonstrate the successful 
application of the test system and the ability of the ACPD to 


detect cracks under in-situ steam generator conditions. 


5.2 Crack Initiation and Growth Measurements 


With the successful completion of the phase one testing, 
the test system was used to obtain crack growth data. Initial 
data was provided from a sample used in phase one. With the 
correlation of the potential drop with the IGSCC growth 
established, the micrographs and data taken from the ACPD were 


used to obtain stress intensity and crack growth measurements. 


Figure 5-13 is the composite micrograph of the fracture 
surface of the sample. The white arcs outline the five 
individual crack fronts. In each case, the fronts are 
approximately elliptical in shape. Measurements of the crack 
front depth and length were performed. Based on the empirical 
relationships developed by J.C. Newman Jr. and I.S. Raju’ for 
this geometry, the stress intensity factors were calculated. 
Crack growth rates were calculated based on the initial rise in 


the 10 kHz ACPD signal for the reference area. 


Figure 5-14 illustrates graphically the approximate 
location of the crack initiation at 540 hours. Results of the 
rack stress intensity and growth rate calculations are 


tabulated in Table 5-1 for the five crack growth fronts. 


2 J.C. Newman Jr. and I.S. Raju, “An Empirical Stress Intensity Factor 


Equation for the Surface Crack,” Engineering Fracture Mechanics, 
КЕ No.l-2 (Great Britian:1981) ,рр. 185-192. 


76 





2 


u dm P К ES 4 
Ductile Fracture ~~ 
| P. 4" р x 


рони È 


1 5 
Е d 
^o a А 
ee 
à T WEY wre © 
- " d 4 js 
M i» ee %; 
2 <, № 
“ = ~ 4 2 


E В 


b .... - № \ 
U Te Ss 85 


vn Outer Surface " 





Figure 5-13: SEM Composite Micrograph of IGSCC Fracture Surface Showing 
Elliptical Crack Fronts. 


10 kHz Signal 


1 kHz Signal 


Normalized Potential Drop (1 kHz) 
(ZH OT) doaq T€rausao0d рэ2ттемаон 





300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 


Time (hours) 


Figure 5-14: Normalized ACPD for Sample Shoving Crack Initiation Time Based 
on Increasing 10kHz Potential Drop. 


Length (2a) 
(mm) 


15.3 207 





Table 5-1: Tabulated Results for Stress Intensity and Grovth Rates. 


79 





Additional crack growth data was obtained for phase two. 
To expedite the crack initiation process, a sample with a well 
defined axial crack initiated by fatigue (as described in 
BSotdon 4) was used for sample 960212. The machined notch was 
elliptical with an estimated depth of 0.5 mm from dial indicator 
readings and measurements conducted from the crack geometry 
(Figure 5-15). The fatigue crack growth front was expected to 
be elliptical, demonstrated in prior testing. Based on this 
geometry, the fatigue crack depth was calculated to be 1 mm or 


B vof the wall. 


Fatigue Growth Area | P mm Notched Area 


\ 


4 4 | 
.- Section A-A' Dt, ——- Sample Tub 


— — Notch in Test Area . | 
With Fatigue Crack Axial Section A-A' 





Figure 5-15: Schematic of Tube Samples With Fatigue Pre-Cracks in Test Area. 


Procedures for system operation outlined in Section 4 were 
followed. The tube sample was polarized to +150 mV versus 
nickel based on the maximum corrosion rates observed by J. 
Mimsden, S.L. Jeanjaquet, J.P.N. Paine апа А.К. McIlree. 


Initial wall stress (38%) was chosen to yield an initial stress 


mJ. Lumsden, $.L. Jeanjaquet, J.P.N. Paine and A.R. Mellree, "Mechanism and 


Effectiveness of Inhibitors for SCC in a Caustic Environment,” Seventh 
International Symposium on Environmental Degradation of Materials in Nuclear 
Power Systems-Water Reactors, Vol.1 (Breckenridge, CO: NACE International, 
ШЕСІ 7-10, 1995). pp. 317-323. 


80 





intensity factor at the crack front of 11 MPa\m (10 ksivin). 
This value was chosen based on IGSCC crack growth rates observed 
in ASTM samples by M. Miglin’. The intent was to prevent fast 
crack growth and penetration of the tube wall. This in turn 
would allow observation of the crack growth by the ACPD system 
as demonstrated during the first phase of testing. Applied wall 
stress and sample temperature as a function of time are shown in 


Figure 5-16. 


325 


320 


315 


310 


Sample Temperature (C) 
(s89338 DT9TA +2'0 30 $) 583129 TTEN 


305 
Sample Temperature 
O wall Stress 





7 
ө 
ө 
o 
ө 





О 50 100 150 200 250 300 350 


Time (hours) 


Figure 5-16: Sample Temperature and Applied Wall Stress (Percent of 0.2% 
Yield Stress). 


BM. Miglin, J.V. Monter, C.S. Wade, M.J. Psaila-Dombrowski, A.R. Mcilree, 


BSCC of Alloy 600 in Complex Caustic Environments,” Seventh International 
Symposium on Environmental Degradation of Materials in Nuclear Power 
Systems-Water Reactors, Vol.1 (Breckenridge, CO: NACE International, August 
10. 1995) pp. 277-290. 


81 





The potential drop measurements versus time are shown in 
Figure 5-17. An upward movement of the test sample caused the 
repositioning of the pickup wires and the large increase in 
potential at 140 hours into the test. At 185 hours into the 
test, with no large increases in the potential drop indicating 
crack growth, the wall stress was increased to 58%, or a K value 
of 16.4 MPa\m (215 ksiVin) based on initial crack geometry. The 
test terminated at 310 hours when a system shutdown occurred due 
to a loss of internal sample pressure. Subsequent examination 
of the test sample indicated that a through wall crack had 


E urred ам пе pre -noteied test site (Figures 5218 and 5-19). 


Analysis of the normalized plot of the potential drop test 
area versus reference shows an overall increase in both the 10 
ЕНА and 1 kHz potential drop signals from the initiation of 
Ес Ето (Figure 5-20). Based on this, the duration of crack 
Erowth was 3410 hours with IGSGCC initiation occurring from the 


Enset of thewtest. 


4. 


| 
‚ № 1 kHz Test Area 
1 
3. 
| 1. 


x 
ЈЕР 


ж 
10|kHz Test Area 


д | 


1 kHz Reference 


Potential Drop (milli-volts) 





Time (hours) 


Figure 5-17: Potential Drop Versus Time for Test and Reference Areas. 


82 





„Тһгоиов-Ман. га 
Crack 





[d n 


Figure 5-18: Stereo Microscope Image of the Through-Wall Crack Seen at the 
Base of the Machined Notch. 





Figure 5-19: Low Power Stereo Microscope Image Showing the Relative Position 
of the Through Wall Crack to the Probes. 


83 





10kHz Signal 


ZHX OT - doaq тетзиззоа paztTwuaon 


Normalized Potential Drop - 1 kHz 


1. ikHz Signal 





0 50 100 150 200 250 300 350 
Time (hours) 


Figure 5-20: Normalized Potential Drop (Test Area with respect to the 
Reference Area) Showing an Increasing Signal from the Onset of Testing. 


The sample tube was sectioned and the crack ductilely 
fractured. A visual examination was conducted by SEM. Figure 
5-21 clearly shows a region of IGSCC. Several regions of 
through-wall crack propagation can also be seen. Figure 5-22 is 
a magnified micrograph of the IGSCC area showing two distinct 
growth areas for the two different applied stress intensities. 

À beach mark separates the two varying areas of IGSCC growth and 
is outlined in white. Micrographs of the ductile region, caused 
by the process used to open the crack area, and the pre-fatigue 
regions are presented in Figures 5-23 and 5-24 for comparison. 
The transition region between the fatigue area and the IGSCC 


Erca іс shown in Figure 5-25. 


84 








Area S owes 
Preure 2-22 


Figure 5-21: SEM Composite Micrograph of Fracture Surface Clearly Showing 
Area of IGSCC. 


E MOIS CAT | 


О 


EXE АКТ ЛА 


гокџ 675% 133Р 0687 





Figure 5-22: Magnified SEM Micrograph of IGSCC Area Outlining Two Possible 
Crack Growth Areas. 


85 








Figure 5-25: Micrograph Showing the Pre-Fatigue to IGSCC Transition Region. 


86 





Table 5-2 summarizes the results of crack depth 
measurements for the two crack growth regions indicated by 
Meare 5-21. Crack growth time is based on initiation of IGSCC 
at the onset of testing from the ACPD data. Based on the visual 
measurements of the fracture surface, the initial stress 
intensity (K,;) of 11.9 МРаУт (10.9 ksiVin) compares favorably 
with the estimated value during the test. The area of IGSCC 
measures 67um and 203um, respectively, in depth from the surface 
for the two growth areas. Finite element analysis yields an 
average Kj, of 12.3 MPa\m and 17.7 MPaYm, respectively, for the 


two growth areas. 


Area Depth (b) Length (2a) К dA/dt 
(ит) (тт) МРаҮт (mm/yr) 
EXFatigue | 1114 | Боо  À 2| 11.9 | ma  —— 


1181 
NNNM ss 


Table 5-2: Tabulated Results for Stress Intensity and Growth Rates. 










87 





6. Discussion 


6.1 ACPD Proof of Principle 


The goal of phase one, to prove that the ACPD system could 
be used to detect cracks on tubing under in-situ steam generator 
conditions, was validated. The data clearly indicate a 
correlation between potential drop and crack initiation and 
growth. This was illustrated in the normalized plots of Figures 
5-3 and 5-6. To substantiate this claim, predicted values for 
the potential drop can be calculated and compared to the actual 
values obtained. The assumption that the 10 kHz and 1 kHz 
signals are close to DC signals will be made for simplicity. 
This assumption is valid since the non-ferromagnetic properties 
of alloy 600 provide little increase in current density near the 


Eurface. 


For one sample, the predominant effect was an area change 
since the current path length to the crack was several orders of 
magnitude greater than the change of length due to the crack. 

The normalized calculated potential drop versus actual results is 


plotted in Figure 6-1. 


In a second sample, the IGSCC crack occurred under the 
reference probe vice test area. This was probably due to higher 
residual stresses in this area due to the weld (a small heat 
affected zone) or incomplete plating that allowed a crevice. The 
difference between the calculated potential drop from the area 
change in the sample and the increase in the path due to crack 
penetration is an order of magnitude. Therefore, the potential 
drop was most affected by the change in the path length. This is 
BEvalid assumption for an AC signal, deviating from the initial 


assumption of pure DC. Using the depth of penetration shown in 


88 





Table 5-1 and assuming a constant linear crack growth, the 
normalized calculated potential drop versus actual test results 


Me plotted in Figure 6-2. 


The correlation between the actual and calculated potential 
drops is excellent. Differences between the actual and 


calculated potential drops can be attributed to: 
gi current loss through the conductive solution, 


(2) relative resistance differences between the wires and 


sample from those assumed, and 


a umnduetdve losses not included in the calculation since DC 


Burrent was assumed for the estimate. 


The 1 kHz signal in one set of data shows a much greater rise 
than the 10 kHz predictions but only near the end of the test 
cycle. However, if one assumes the crack started at the same 
point the 10 kHz signal began to rise, the overall increase in 
the 1 kHz signal is also very close to that calculated. The 
possibility of noise in the 1 kHz signal during the 520 to 1400 
Bur region could account for the failure of the potential drop 
to rise in this period. This postulate became more likely when 
it was discovered that a ground in the shield for the 1 kHz 


signal existed. 


89 





Normalized Potential Drop (1 kHz) 


Normalized Potential Drop (Test Area/Reference) 





Normalized Potential Drop for Test 950909 


1.40 
1.35 
1.30 
1.25 
1.20 
1.15 
1.10 
1.05 


Calculated Potential Drop 


.- 


1.00 
0.95 
0.90 
0.85 
0.80 
0.75 
0.70 Actual Potential Drop From Test 950909 
0.65 
0.60 
0.55 
0.50 
0.45 
0.40 

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 


Time (hours) 


Figure 6-1: Calculated Potential Drop Versus Actual. 


Normalized Potential Drop for Test 950909C 
2.000 


1.625 


Az a 
- 
-. 
ES P 


ж 


"mia 


= 
> 


- 5 5 
ж ғ 


u 
áo seo P 
- d 10 kHz Signal 1.250 


- 
Tuum ea ' 


1 kHz Signal 0.875 


0.500 
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 


Time (hours) 


Figure 6-2: Calculated Potential Drop Versus Actual. 


90 





({2Н бот) dozq Tetausaog paztTeuzoy 





6.2 Crack Growth Rates 


The measurements of crack length and depth given in Tables 
5-1 and 5-2 were used to calculate stress intensity and growth 
rate based on test duration. The average values of stress 
intensity and calculated growth rates for the two sets of data 


for phase two are plotted in Figure 6-3. 


= 
o 


Crack Growth Rate (mm/r: 


= 


Stress Intensity (Ki) (MPa root m) 





Figure 6-3: Alloy 600 Tube Crack Grovth versus Stress Intensity from Phase Two 
Testing. The Shaded Area Represents Data Ranges for K and Grovth Rates from 
M. Miglin’ : 


It is significant to note an apparent relationship in the 
Stress intensity versus crack growth rates. Based on the tube 


data collected, growth rate appears to be exponential to the 


Км. Miglin, J. V. Monter, C. S. Wade, M. J. Psaila-Dombrowski, A. R. Mellree, 


”SCC of Alloy 600 in Complex Caustic Environments,” Seventh International 
Symposium on Environmental Degradation of Materials in Nuclear Power Systems- 
Water Reactors, Vol. 1 (Breckenridge, CO:NACE International, August 7-10, 

Ш 005), рр. 277-290. 


91 





applied stress intensity when K, is less than about 15 MPa\m. 
For values between 15 and 20 MPaYm, crack growth rates 
accelerate, indicating a threshold value, approximately 15 MPa\m, 


above which growth rates increase at an increasing rate. 


Previously reported crack growth data in 10% NaOH was 
compared with the results from this study. M. Miglin, et.al., 
have reported stress intensity values from 5 to 40 MPaYm with 
growth rates from 0.2 to 76 mm/yr’. The shaded area in Figure 6-3 
represents this broad range of reported data from M. Miglin. In 
addition, Rebak, et.al. have obtained crack growth data for Alloy 
600 flat tubes at various pH and temperatures ranging from 300 to 
680907. А good correlation is shown when comparing Rebak's data 
to the present tube data (Figure 6-4). The dark upper bands 
represent data primarily in pH range 9 to 10, while the lower, 
lighter band represents data primarily in the 5 to 9 pH range. 
The data indicate a crack growth rate dependance on pH as well as 


on Ky . 


Figure 6-5 shows a plot of various growth rates versus pH 
from Miglin’ and Rebak, et.al.*. The data from Rebak is broken 


down into stress intensity ranges and is consistent with the 


? M. Miglin, J. V. Monter, C.S. Wade, M. J. Psaila-Dombrowski, A. R. McIlree, 
"SCC of Alloy 600 in Complex Caustic Environments," Seventh International 
Symposium on Environmental Degradation of Materials in Nuclear Power Systems- 
Water Reactors, Vol. 1 (Breckenridge, CO: NACE International, August 7-10, 
095), рр. 277-290. 
* T. Beineke, et. al., "PWR Molar Ratio Control Applications Guidelines", EPRI 
ЖЕ ЕЕ 10%8511-У1,1995, р. 2-13. 
* К.В. Rebak, and Z. Szlarska-Smialwska, "Influence of Stress Intensity and 
Loading Mode on Intergranular Stress Corrosion Cracking of Alloy 600 in 
Primary Water of Pressurized Water Reactors," Corrosion, Vol. 50, No. 5, (May 
1994), ыры 378-393, 

Миа. etal. гор. cit., p. 285. 
* R.B. Rebak, and Z. Szlarska-Smialwska, “Influence of Stress Intensity and 
Loading Mode on Intergranular Stress Corrosion Cracking of Alloy 600 in 
Primary Water of Pressurized Water Reactors," Corrosion, Vol. 50, No. 5, (May 
№994), pp. 386-387. 


92 





results illustrated in Figure 6-4 (Note the decrease in growth 
rate with pH for a specific stress intensity band). The crack 
growth data for prior testing is based on ASTM specimens, not 
tubes. Of significance is the wide variation in growth rates in 
pH range 10 (.025 to 250 mm/yr.). The alloy 600 tube data 


obtained in phase two fall in the middle of this range. 


ој | 


о Керак - рн 5-6 

u Керак - рн 6-7 

2 Керак - рн 7-8 

m Rebak - pH 8 

e Present Test Data - pH = 10 


— 
o 
o 


: || 
HIS ae 
| | HIPS] 
ТН ГМ 


5 
E 
E 
© 
3 
£ 
$ 
e 
o 
5 
5 


и“ F |o F 
E 
HAN 


EUN 
E eae 5-8 Band 


Stress Intensity (MPa root m) 





Figure 6-4: Comparison of Crack Growth Data versus Stress Intensity (K). 


93 





D 
Г) 
| 
> $ 
Qo 
tod z 


Rebak K=80-101 MPa root m 


Rebak K= 20-80 MPa root m 
pm rtm a 


\ 


è 


e Present Test Data (K = 4 - 18 MPa root m) 
5 M. Mglan Data ( K =5 - 40 МРа год! т) 

4; Rebak Data(K = 0 - 20 MParoot т) 

© Rebak Data(K = 20 - 40 MPa root m) 

x Rebak Data(K = 40 - 80 MPa root m) 

о Керак Data (K = 80 - 101 MPa root m) 


| х 
t> 


& 
E 
E 
Ф 
3 
Е 
Ұ 
2 
o 
x 
5 
о 


о 
о 
> 


рн @ 315С 





Figure 6-5: Comparison of Alloy 600 Crack Grovth Rates versus pH. 


No prior data exists for crack growth in actual steam 
generator tube samples. The tube data presented here represents 
the first in-situ monitoring and crack growth measurement of 
steam generator tubing. The tubing data obtained during phase 
E c testing correlates with data reported for non-tube (i.e. 
plate samples) alloy 600 specimens, which implies credibility for 
the results from plate samples. Based on this data and the 
relationships obtained, steam generator tube crack growth rates 
from in-situ inspections can be predicted and further development 


of an on-line monitoring system using ACPD methods developed. 


94 





E. Conclusions 


A laboratory system capable of monitoring alloy 600 steam 
generator tubing for IGSCC using an ACPD multi-frequency 
detection system was constructed. The system provided the 
capability to maintain an environment similar to an operating 
steam generator secondary side. Additionally, the autoclave 
system provided the capability to impress varying wall stresses 
on the tube samples up to 140 % of yield strength to expedite 


аск initiation. 


The capability of the ACPD system to detect initiation and 


growth of IGSCC on alloy 600 tubes in the in-situ environment was 


demonstrated.  IGSCC initiation and growth was monitored up to 
1240 hours at two frequencies. The higher frequency proved to 
have a greater sensitivity to crack initiation. The measured 


potential drop from the tube data correlated well with calculated 


values. 


IGSCC crack growth data for alloy 600 tubes was obtained. 
Average values of stress intensity (Kı) ranged from 4.2 to 17.4 
MPa\m with Prowtheratesstrom 1.6 to 12.3 mm/yr. Data obtained 
from these tests show an exponential relation between crack 


growth rate and stress intensity for alloy 600 tubes. 


The crack growth data obtained from these tests were 
compared to data from previous researchers using standard test 
samples of alloy 600. The measured tube crack growth data and 
stress intensities compared favorably to prior data. The data 
obtained is the first IGSCC crack growth data in tubes. This 
data and future test data obtained using the procedures and 
techniques developed, may be used in predicting tube failure 


times based on site inspections. 


95 





8. Recommendations for Future Work 


Àn early goal of the program was to make use of the ACPD 
monitoring system to measure the effectiveness of various 
inhibitor compounds to mitigate or halt SCC growth. Due to 
ше constraints, this could not be accomplished in this 
program. However, inhibitors can play a critical role in the 
mitigation of IGSCC in existing steam generators. The system 
developed in this research can provide the data essential for 
an evaluation of inhibitors and this evaluation should be 


performed. 


С Кола Stress intensity (K) versus growth rate data is 
needed to better establish the relationship between growth rate 
апа К. This data will also establish a better band of expected 
prowth rates for alloy 600 tubing. 


Data from ACPD measurements of SCC growth in tubes is 
required to provide a predictive tool. Future data collection 
will enable the determination of crack size versus potential 


drop and allow operator prediction of crack size on-line. 


Finally, higher frequency application of the ACPD input 
signal needs to be investigated to allow the probe spacing to 
increase with no loss in sensitivity. The use of the ACPD 
system under actual steam generator conditions will require 
less restrictive placement of the output probes. The data 
collected during testing conducted thus far indicates that 
higher frequency application may allow an increase in probe 


spacing with the same sensitivity to crack growth. 


96 





References 


Adams, James P. and Eric S. Peterson. “Steam Generator 
Secondary pH During A Steam Generator Tube Rupture.” Nuclear 
meehnology, CII, (June 1993). рр. 304-306. 

Andersen, P.L.. "Effects of Temperature on Crack Growth Rate in 


Sensitized Type 304 Stainless Steel and Alloy 600.” Corrosion 
Science. Vol.49, No.9 (September 1993). 


Babcock and Wilcox. Steam/its generation and use. 39th Ed. New 
York: Babcock and Wilcox, 1978. 


Ballinger, R.G. and I.S. Hwang. "Characterization of 
Microstructure and IGSCC of Alloy 600 Steam Generator 
паше. Einal Report, EPRI TR-101983, Palo Alto: Electric 
Power Research Institute (February 1993), p. 3-56. 


Bandy, R., R. Roberge, and D. van Rooyen. ^Intergranular 
Failures of Alloy 600 in High Temperature Caustic 
Preironments. Corrosion - NACE”, XVI no. III, (March 1985), 
poe 142-151. 


Balmeleter, T.. E. Avallone Т. Baumeister III. Mark’s Standard 
Handbook for Mechanical Engineers. 8th Ed. New York: McGraw- 
Hill Book Co. (1978). 


Beineke, T.. et.al..”PWR Molar Ratio Control Applications 
mude lines”. EPRI TR-10481 -V1,1995. 


Berry, Warren E. Corrosion in Nuclear Applications. New York: 
John Wiley & Sons, 1971. 


и в В.С. Lord, T.A. Prater and L.F. Coffin." The 
Reversing DC Electrical Potential Method. Schenectady, NY: 
General Electric. 


ес ЈЕ... Caustic Stress Corrosion Cracking Studies at 
288C(550F) Using the Straining Electrode Technique. 
Corrosion NACE, V34, No.6 (June 1978). 


Charlesworth, F.D.W. and W.D. Dover. “Some Aspects of Crack 
Detection and Sizing using A.C. Field Measurements.” The 
Measurement of Crack Length and Shape During Fracture and 
Fatigue. Edited by C.J. Beevers. West Midlands,UK:EMAS 
(1982), p. 258. 


97 





Electric Power Research Institute. "Boric Acid Application 
Guidelines for Intergranular Corrosion Inhibition," EPRI NP 
E558. (1984). 


Electric Power Research Institute. Steam generator Reference 
Handbook. EPRI TR-103824, Project 2895;4044 Nuclear Power 
Encubp. Palo Alto: EPRI, Dec 1994. 


Friend, Wayne Z. Corrosion of Nickel and Nickel Based Alloys. 
New York: John Wiley & Sons, 1980. 


Hwang, I.S. and R.G. Ballinger. "A Multi-Frequency AC Potential 
Drop Technique for the Detection of Small Cracks." 
Measurement Science Technology, Vol 3,(1992), p 63. 


Hwang, I.S., "Embrittlement Mechanisms of Nickel-Base Alloys in 
Water." Ph.D. dissertation, Department of Nuclear 
Engineering, Massachusetts Institute of Technology, 1987. 


Inco Alloys International. Inconel. Huntington, WV:Inco Alloys 
International, Inc. 


International Nickel Co., Inc.. "Corrosion Resistance of Nickel 
and Nickel-Containing Alloys in Caustic Soda and other 
Pukalies.” Corrosion Engineering Bulletin. CEB-2, 1971. 


Jang, I., “Effect of Sulphate and Chloride ions on the Crevice 
Chemistry and Stress Corrosion Cracking of Alloy 600 in High 
Temperature Aqueous Solutions.” Corrosion Science , Vol. 


33, No. 1 (1992), pp. 25-38. 


Johnk, Carl T.A. Engineering Electromagnetic Fields and Waves, 
New York: John Wiley and Sons, 1975. 


Bohns, D.R. and F.R. Beckitt. “Factors Influencing the Thermal 
Stabilisation of Alloy 600 Tubing Against Intergranular 
me osion. Corrosion Science, XXX No. ІТ/ІІТ (1990), pp. 
29-237. 


Jordan, Edward C. Electromagnetic Waves and Radiating Systems. 
POdSEd. Englewood Cliffs, N.J.:Prentice-Hall, 1968. 


King, Ronold W. P. Fundamental Electromagnetic Theory. 2nd Ed. 
New York: Dover Publications, Ine, 1963. 


98 





Eumsden, J. B., S.L. Jeanjaquet, J.P.N. Paine and A. Mcllree. 
“Mechanism and Effectiveness of Inhibitors for SCC in a 
Caustic Environment.” Seventh International Symposium on 
Environmental Degradation of Materials in Nuclear Power 
Systems - Water Reactors, Vol 1, Breckenridge, CO: NACE 
Maeerhnational (August 7-10,1995), pp. 317-325. 


EN den. J.B. and P.J. Stocker. *Inhibition of ICA in Nickel 
КЫ = ШАГ сув in Caustic Solutions." Corrosion/88, Houston, 
TX: National Association of Corrosion Engineers (1988). 


ШІ, M.T., J.V. Monter, C.S. Wade, M.J. Psaila-Dombrowski, 
sad A,R, Mellree. “SCC of Alloy 600 in Complex Caustic 
Environments.” Seventh International Symposium on 
Environmental Degradation of Materials in Nuclear Power 
Systems - Water Reactors, Vol 1, Breckenridge, CO: NACE 
Murcrnational (August 7-10,1995), pp. 277-290. 


Morra, M.M., J.M. Morra and R.R. Biederman. "A Technique for 
the Preparation of Powders for Examination by Transmission 
Electron Microscopy.” Materials Science and Engineering, A124 
(#1550) pp. 55-64. 


Шелд Је., J.C. апа I.S. Raju, “An Empirical Stress Intensity 
Factor Equation for the Surface Crack.” Engineering Fracture 
КО О а тсс, Vol 15, No. 1-2 (Great Britain:1981), pp. 185-192, 


Payne, M. and P. McIntyre. “Influence of Grain Boundary 
Microstructure on the Susceptibility of Alloy 600 to 
Intergranular Attack and Stress Corrosion Cracking.” 
Werrosion-NACE, XLIV no. V, (May 1988), рр. 314-319, 


Rebak, R.B. and Z. Szlarska-Smialwska, “Influence of Stress 
Intensity and Loading Mode on Intergranular Stress Corrosion 
Cracking of Alloy 600 in Primary Water of Pressurized Water 
КС сога, Corrosion, Vol. 50, No. 5, (May 1994). pp: 3278- 
393. 


Bios, R., T. Magnin, D. Noel, O. DeBouvier. “Critical Analysis 
of Alloy 600 Stress Corrosion Cracking Mechanisms in Primary 
Water." Metallurgical and Materials Transactions A, Vol. 26, 
Mem 4, (1995), pp. 925-939. 


Roberts, J.T. Adrian. Structural Materials in Nuclear Power 
Systems. New York: Plenum Press (1981). 


Ecaffer, Saxena, Antolovich, Sanders and Warner. The Science 
and Design of Engineering Materials. Chicago: Irwin (1995). 


99 





Scarberry, R.. "Corrosion of Nickel Based Alloys." Conference 
Proceedings. American Society of Metals. Metals Park, 1985. 


в М, ЮВ. Lowensteln, A.P.L. Turner, S.R. Ward. J.A. 


Gorman, P.E. MacDonald, G.H. Weidenhamer. “Assessment of 
Primary Water Stress Corrosion Cracking of PWR Steam 
Generator Tubes.” Nuclear Engineering and Design, Vol. 134, 


No. 2-3, pp. 199-216. 


Ehen, Y. and P. G. Shewmon. “Intergranular Stress Corrosion 
Cracking of Alloy 600 and X-750 in High Temperature Deaerated 
Meeer/Steam.” Metallurgical Transactions A, Vol. 22A, No. 8, 


ШЕЕ Е 1991), pp. 1857-1864. 


ООСО К J. Koch, T. Angeliu, and G.S. Was. “The Effect of 
Grain Boundary Chemistry on Intergranular Stress Corrosion 
Cracking of Ni-Cr-Fe Alloys in 50 Pct NaOH at 140°C.” 
Metallurgical Transactions A, Vol. 23A, No. 10, (October 
052) pp. 2887-2904. 


Todreas, Neil E. and Mujid S. Kazimi. Nuclear Systems I. 
ой, PA: Taylor € Francis, 1990. 


Uhlig, Herbert H. and R. Winston Revie. Corrosion and Corrosion 
Control. New York: John Wiley & Sons, 1985. 


Van Vlack, L. H.. Elements of Material Science and Engineering. 
Оке. Reading, MA: Addison-Wesley Publishing Co. (1975). 


Verpoest, I., E. Aernoudt, A. Deruyttere, and M. Neyrinck. "An 
Improved A.C. Potential Drop Method For Detecting Surface 
Microcracks During Fatigue Tests of Unnotched Specimens." 
Fatigue of Engineering Materials and Structures, Vol 3 
ВИС), вр. 200-209, 


Wagner, C.. Discussions at the First International Symposium on 
Passivity, Heiligenberg, West Germany, 1957. Corrosion 
Demente. Vol 5, (1965). 


Woodward, J. “Rapid Identification of Conditions Causing 
intergranular Corrosion or Intergranular Stress Corrosion 
Cracking in Sensitized Alloy 600.” Corrosion - NACE, XL, No. 
XII, (December 1984), pp. 640-643. 


100 





À-l. Autoclave System Operating Characteristics 


This appendix contains information describing heater 
controller settings, autoclave heatup characteristics, and a 
description of the autoclave head ring failures which occurred 
during early testing due to a failure to completely purge the 
autoclave volume of oxygen. This information is provided to 
provide a benchmark for system operating performance during 


future testing. 


A-1.1 Omega CN9000A Heater Controller Parameter Settings 


The CN9000 Heater Controller manufactured by Omega 
Engineering, Inc. is a Proportional-Integral-Differential 
controller. The controller has a self-tune mode which allows 
the parameters to be automatically calculated during the 
heatup. However, the auto-tune feature has a time-out routine 
in the software which will terminate the auto-tune calculation 
after a predetermined time, which is preset by the 
manufacturer. This occurred during the heatup tests in the 
static mode. As an alternative, the PID parameters may be 
manually calculated from observation of the autoclave 
temperature oscillations during heatup. Detailed procedures 


are in the CN9000 Operators Manual (M575/0289). 


A heatup of the autoclave was conducted with a set 
temperature of 280°C in order to obtain the necessary 
parameters without exceeding the 315°C rated autoclave 
temperature. When the set temperature was reached by allowing 
the controller to operate in a manual mode with the PID 
controller inactive, oscillations occurred as expected. 
Measurements of the amplitude and period of the oscillations 


were obtained from a graph of the temperature versus time from 


Oa 





the LABVIEW® (figure A-1). The required parameters are 


calculated and presented in Table A-1. 


325 


315 
305 
295 
285 
275 
265 
0 10 20 30 40 50 60 70 80 90 100 


Time (m) 


Autoclave Temperature (C) 


Figure A-1-1: Heatup 950726 Showing Oscillations of the Heater Controller. 
The Amplitude Was Measured at 9°C with a Period of 15 Minutes. 


CN9000A Heater Controller PID Parameters 


Measured Amplitude: 9°C Measured Period: 15 
minutes 


PID Parameter Governing Equation Result 
(Setting) 


T/20 


(%) 
seconds 
seconds 
Table A-1-1 : CN9000A PID Controller Parameters for Static Autoclave 
Operation 








102 





А-1.2 Heatup Plots 


Time 
autoclave 
generates 
operating 
autoclave 


a greater 


Біліп and 


to heatúup isa function of the fluid quantity im the 
and the heat loss of the system. The heating mantle 
2700 watts total heat output. Based on the two 
modes, this output is sufficient to heat the 

to the desired 315°C. However, since heat is lost at 
rate in recirculate mode due to the incoming cold 


larger fluid volume in the autoclave, additional 


insulation on the head and preheating of the incoming fluid is 


necessary 


to provide reasonable heatup times. Testing showed 


that heatup to temperature requires approximately two hours in 


static mode and six hours in recirculate mode. Figure A-2 


shows the heatup plots during normal static and recirculate 


heatups. 


The fluid is preheated to 80°C using a heat tape on 


the incoming tubing for the recirculate mode of operation. 


Temperature (C) 





Static Heatup 


а. 
(eise ta mtn 
u 

2... 


Recirculation Heatup 


100 200 300 400 


Time (minutes) 


Figure A-1-2: Heatup Plot Showing Contrast in Time Between Recirculation 


and Static Modes 


193 





A-1.3 Autoclave Head Seal Ring Failures 


Cracking of the nickel O-ring which seals the head and 
autoclave body occurred during two early caustic tests while 
operating in recirculation mode. The failures occurred within 
100 hours of reaching final autoclave temperature. Post 
shutdown inspections revealed a green residue on the interiors 
of the autoclave head and body. Although no formal chemical 
analysis was performed, the residue was thought to be nickel 
oxide or nickel hydroxide. Inspection of the O-ring indicated 


radial cracking with signs of erosion in one location (figures 


БЕ бапа А-л). 





a 
Бк 





Figure À-1-4: Erosion Damage to O- 


Figure A-1-3: Channel in O-ring и 
from Test 950703 Ring from Test 950403 


Based on the best available data from the post-shutdown 
analysis, the cracking is thought to be from oxygen which was 
Hot purged from the nipples prior to heatup. The oxygen was 
forced into solution during the subsequent fill. Once at 


temperature, the oxygen in caustic attacked the O-ring which is 


104 





subject to a large compressive stress. The greater thermal 
expansion of the nickel versus the carbon steel head bolts 
further increases the compressive stress on the O-ring during 
heatup. Areas of increased stress due to a prior flaw are 
probably where the cracking initiated. The failure of the O- 
rings resulted in system shutdown due to a loss of autoclave 


pressure. 


Future O-ring cracking while operating in recirculation 
mode should not occur. The autoclave head was modified to 
allow thorough purging of the oxygen in the nipples by 


iux dlatyon of valves at the high points (ACV-1 through 6). 


105 





A-2. Operating Procedures 


This appendix contains detailed procedures for operating 
the autoclave system in all modes. Emergency procedures are 
included to rapidly shut down the system in an emergency. 
Operation of the system using these procedures should prevent 


equipment damage and minimize any personnel hazards. 


106 





1 D M 

hold indy (92 кмм А Богі; 
ата ветућђегола ходају зена „ason. ine 
„тој или ли 25 mor nye adi лов 3 P "x 


271017934 hiwede *o34ube^soTq вәшіз aatan en T Ы 
аБзалай іғайок?ың таз oximtziw b P 







йй 


A-2.1 System Startup Procedure - Static Operation. 


A-2.1.1 Prerequisites 


a. The sample is installed in the head and the head is 
bolted onto the autoclave. Ensure the head is torqued to 125 
to 140 foot-pounds. Verify all Conax fittings are torqued to 
the required values to prevent hydrogen leakage. 


b. All external tubing connections are installed to the 
head. The heating mantle is installed with required thermal 
Пе Л оп bolted in place (i-e. lower autoclave thermal 
Bucsulation disk). 


c. Cooling jacket water is connected to the head. Ensure 
cooling water is turned on and flow verified prior to startup. 


d. The main electrical panel breakers are shut for 250 
volt service to the heater controller. 120 volt service is 
available to the heater controller and control systems. Verify 
all heating mantle electrical connections are complete. 


e. 50 to 75 psig compressed air service is connected to 
regulator REG-3 and solenoid S-2. 


f. Nitrogen bottle pressure should be a minimum of 1500 
psig to prevent frequent booster pump cycling. 


g. Hydrogen bottle pressure is a minimum of 2000 psig. 
Pressure below this will result in operation closer to solution 
saturation temperature. 


bee solenoid 5-2 Emergency Dump Switeh is in “ӘНШІ”. 
i. System Override Switch is in “STARTUP”. 


j. The system has been purged of oxygen by an inert gas 
for at least 15 minutes through each valve ACV-1 through ACV-6 
(NOTE: if the gas inlet is through the outlet union cross, 
then purging through ACV-2 is not required). 


k. The injection system is connected or capped at the 
autoclave standoff. 


ir. Connect or verify connected all thermocouple leads. 
This includes the dual junction thermocouple for internal 
autoclave temperature and the return tubing external 
temperature. 


107 





m. The following valve lineup is completed: 


Static System Operation : Startup Valve Lineup 


О-о 5 
5 


АС 


АС 


А 


ОРЕМ 


Table A-2-1: Static System Start-Up Valve Lineup 


В 
[2 
P-4 

| 





re 271.2 Procedure 


a. SLOWLY open the Hydrogen Tank Isolation Valve to 
pressurize the system up to the inlet of REG-2. 


b. Operate REG-2 slowly to pressurize the system to 1800 
psig. Adjust REG-2 as necessary to maintain the system 
pressure 1750 to 1850 psig. 


c. Once the system pressure is 1750 to 1850 psig, the 
sample may be internally pressurized to obtain the desired 
stress across the wall. See the sample pressurization 
procedure. When the sample internal pressure exceeds the low 
pressure shutdown setpoint, reset the shutdown on meter NP-1. 


d. Reset meter ACP-1 once pressure exceeds 1600 psig. 


108 





e. When all shutdown signals are reset, place the System 
Override Switch in “NORMAL”. 


Ber or verify the setpoin: temperature ds 31552 C on 
the CN9000 Series Heater Controller (See the Operators Manual). 


g. Turn the CN9000 Series Heater Controller on per the 
Operator’s Manual instructions. Shut the heater latch 
contactor on the rear of the control panel by pushing the green 
“START” button. A distinct latch sound should be heard. If 
not, check for a power interruption to the heater controller or 
тко! circuit. 


h. Monitor the heatup to verify normal heatup 
characteristics (see Appendix 1). It normally requires 
approximately 2 hours to reach set temperature. 





8-2.2 System Startup Procedure - Recirculation. 


B-2.2.1 Prerequisites 


a. The sample is installed in the head and the head is 
bolted onto the autoclave. Ensure the head is torqued to 125 to 
140 foot-pounds. Verify all Conax fittings are torqued to the 
required values to prevent caustic leakage. 


b. All external tubing connections are installed to the 
head. The heating mantle is installed with required thermal 
insulation bolted in place (i.e. lower autoclave thermal 
Enculation disk). 


c. Cooling jacket water is connected to the head. Ensure 
cooling water is turned on and flow verified prior to startup. 


d. The main electrical panel breakers are shut for 250 volt 
service to the heater controller. 120 volt service is available 
to the heater controller and control systems. Verify all heating 
mantle electrical connections are complete. 


e. 50 to 75 psig compressed air service is connected to 
Fegulator REG-3 and solenoid 5-2. 


f. Nitrogen bottle pressure should be a minimum of 1500 psig 
to prevent frequent booster pump cycling. 


g. The storage tank contains a minimum of 50 liters of cold 
caustic solution which has been adequately deaerated. 


he solenoid 5-2 Emergency Dump Switch is in “SHUT: 
i. System Override Switch is in “STARTUP”. 


j. The system has been purged of oxygen by an inert gas. 
This should be accomplished by backfilling the system with an 
inert gas prior to filling the system solid with caustic. ENSURE 
THE STANDOFFS ARE PURGED VIA ACV-1 THROUGH ACV-6. 


k. The injection system is connected or capped at the 
autoclave standoff. 


l. Connect or verify connected all thermocouple leads. This 


includes the dual junction thermocouple for internal autoclave 
temperature and the return tubing external temperature. 


110 





m. Adjust the recirculation pump flow to maximum (see 
operating manual). 


n. The following valve lineup is completed: 


Recirculation System Operation Startup Valve Lineup 


ACV-1 OPEN THROTTLED 
OPEN 
SHUT 


Valve 


Valve Poss tion 


> 


SHUT 


> 


SHUT 


> 


SHUT 


> 


ШШ 


SHUT 


> 


OPEN 


OPEN 






un 

ns) 
I 

н- 





АСУ-2 КЕСЕ FULL OPEN 


ACV -3 SHUT REG -3 
PNIS ТО VENT 
THROTTLED OPEN 


Table A-2-2: Recirculation System Start-Up Valve Lineup 


SHUT 


OPEN 





2.2.2 Procedure 


a. Place the recirculation pump in “RUN” 


Боасе the system is verified solid (flow sounds) to tank., 
Small pressure increase on meter, etc.), adjust REG-1 slowly to 
pressurize the system to 1800450 psig. 


c. Once the system pressure is 1750 to 1850 psig, the sample 
may be internally pressurized to obtain the desired stress across 
the wall. See the sample pressurization procedure. When the 
sample internal pressure exceeds the low pressure shutdown 
setpoint, reset the shutdown on meter NP-1. 


all 





d. Reset meter ACP-1 once pressure exceeds 1600 psig. 


e. When all shutdown signals are reset, place the System 
Override Switch in "NORMAL". 


ПИ = Бок ерту тре setpoint temperature is 315452 C on the 
CN9000 Series Heater Controller (See the Operators Manual). 


g. Turn the CN9000 Series Heater Controller on per the 
Operator's Manual instructions. Shut the heater latch contactor 
on the rear of the control panel by pushing the green "START" 
Button. A distinct latch sound should be heard. If not, check 
for a power interruption to the heater controller or control 
E rcuit. 


h. Monitor the heatup to verify normal heatup 
characteristics. 





A-2.3 System Shutdown Procedure - Static Operation 
A-2.3.1 Prerequisites 


a. The system is operating at set temperature and pressure. 
No shutdown signals are present. 


ELO 73-2 Procedure 
a. Place the System Override Switch in "OVERRIDE". 
b. Shut the Hydrogen Bottle Isolation Valve. 


c. Reduce the internal sample pressure to atmospheric by 
shutting the Nitrogen bottle isolation valve and reducing air 
pressure to the booster pump to 0 (КЕС-3). Throttle open SP-1 to 
vent the nitrogen to the atmosphere. 


d. Turn the heater controller off (see Operator's Manual). 


e. Open the heater latch contactor by depressing the red 
NS TOP” button on the contactor housing. 


f. If a quick return to operating status is intended, AC-5 
and AC-6 may be shut to maintain hydrogen pressure in the 
scrubber. The autoclave may be vented by throttling open ACV-1 
through ACV-6 or by opening AC-4 to the storage tank. 





A-2.4 System Shutdown Procedure - Recirculation 


A-2.4.1 Prerequisites 


a. The system is operating at set temperature and pressure. 
No shutdown signals are present. 


A-2.4.2 Procedure 
a. Place the System Override Switch in “OVERRIDE”. 


b. Reduce the internal sample pressure to atmospheric by 
shutting the Nitrogen bottle isolation valve and reducing the air 
pressure to the booster pump to 0 (ВЕС-3). Throttle open SP-1 to 
vent the nitrogen to the atmosphere. 


e. Turn the heater controller off (see Operator’s Manual). 


d. Open the heater latch contactor by depressing the red 
ШОО Ор button on the latch contactor housing. 


f. Continue to operate the recirculation pump throughout the 
cooldown. NOTE: A RAPID DECREASE IN PRESSURE MAY OCCUR DUE TO 
SYSTEM CONTRACTION ONCE THE HEATER IS TURNED OFF. This is not 
unusual and should not be mistaken for a pump or backpressure 
regulator failure. Continue to operate the pump and pressure 
will be restored when the makeup exceeds the contraction (at 
about 200°C). 


hen temperature is reduced below 100 C, сре сусвет 


pressure may be reduced by operating REG-1 and the recirculation 
pump secured. 


114 





A-2.5 Emergency Shutdown Procedures 


These procedures may be used to rapidly shutdown the system 
in an emergency (such as a leak) or to place the system in a 
Pautdewhn configuration if an automatic shutdown occurs. 


AS2.5.1 Static System Emergency Shutdown 


a. Open the heater latch contactor by depressing the red 
E OB. button on the contactor housing. 


PROP lace Solenoid 5-2 Emergency Dump Switch in "OPEN". 


c. Shut the isolation valves on the Nitrogen and Hydrogen 
bottles as quickly as possible. 


A-2.5.2 Recirculation System Emergency Shutdown 


a. Open the heater latch contactor by depressing the red 
Eu P button on the contactor housing. 


[Place Solenoid S-2 Emergency Dump Switch in "DUMP". 


c. Shut the isolation valves on the Nitrogen bottle as 
quickly as possible. 


d. If a leak has occurred, secure the recirculation pump to 
avoid additional caustic from entering the autoclave. 
Depressurize the system by rapidly opening REG-1. CAUTION: THE 
SYSTEM MAY BE DEPRESSURIZED BY OPENING AC-2 HOWEVER, CAUSTIC WILL 
VENT TO THE ATMOSPHERE! 





A-2.6 Sample Pressurization System Operation 


EL2-6.1 Startup 
a. Prerequisites 


1. All tubing is assembled to the sample in accordance with 
the manufacturer’s instructions. 


Ze Natrogen bottle pressure is a minimum of 1500 psig to 
prevent frequent cycling of the booster pump. 


3. 50 to 75 psig compressed air service to regulator REG-3 
and S-2 is available. 


4. The system low pressure shutdown is reset or the System 
Bverride Switch is in "OVERRIDE". 


b. Procedure 


1. Slowly open the Nitrogen bottle Isolation Valve. Sample 
pressure on NP-1 and NP-2 should indicate approximately the same 
as the Nitrogen bottle pressure. 


2. Raise the internal sample pressure by operating REC-3. 
The approximate internal pressure as read at NP-1 and NP-2 should 
be 150 times the applied air pressure as read at REG-3. The 
Booster pump will cycle to maintain this pressure within an 
approximate 200 psig band, Allow pressure to stabilize before 
REG-3 is locked to the final setting. 


ING 





A-2.7 Injection System Operation 


А. Prerequisites 


1. Ап inert gas supply is available to deaerate and agitate 
the inhibitor while in the tank. The gas is also required to 
provide a positive head to the injection pump. The gas supply 
line is attached to the fitting on the injection tank. 


2. The solution to be injected is in the injection tank. 
The fill plug on the tank has been replaced. 


3. The following valves are in the position noted: 


ЕЕ | _ 


Table A-2-3: Injection System Start-Up Valve Lineup 









4. 120 VAC Service Power is connected to the injection 
pump. 


5 lf agitation or deaeration of the solution is required, 
attach a vent line to INJ-2 prior to opening. 


6. A collection bottle should be placed at the common vent 
prior to pressurizing the tank. 


7. CAUTION: NEVER START THE PUMP WITH A HEAD GREATER THAN 
000 PSIG AGAINST THE PUMP OUTLET. 


B. Procedure 


1. Pressurize the injection tank to approximately 10 psig 
by adjusting the gas bottle regulator. If deaeration or 
agitation of the solution is required, INJ-2 should be throttled 
open to adjust the flowrate through the injection tank while 
maintaining tank pressure at approximately 10 psig. 


2. Set the injection pump flow to the desired rate with the 
micrometer on the pump. 


П 





Bee start the pump. Verify flow through the vent. After a 
few seconds, place INJ-1 in the “INJECT” position. Monitor 
autoclave pressure carefully to ensure the system does not 
overpressure. 


4. When solution injection is complete, place INJ-1 in the 
ETO VENT" position. Stop the pump. 


5. Shut off the gas pressure to the injection tank. 


6. Vent the tank by opening INJ-2 fully. 





A-2.8 Electroless Nickel Plating Technique 


This electroless nickel plating technique was developed by 
Dr. Martin Morra! as an aid in preserving surface features during 
metallographic preparation. While commercial kits for 
electroless nickel plating are available, this technique was 
developed specifically for superior retention of fine surface 
details. The hard, resilient plating also makes this technique 


ideal for the Alloy 600 sample preparation. 
A=2.8.1 Preparation 


a. Clean all glassware to be used to store or hold the 
plating solution with a 50% nitric acid-50% deionized water (by 
Bolume) solution heated to 90° to 100°C. 


b. Prepare 2 liters of plating solution by mixing 54 grams 
of Sodium Hypophosphite (NaH,PO,eH,0), 40 grams of Nickel Sulfate 
Hexahydrate (NiSO,°6H,O) and 32 grams of Sodium Succinate 
Hexahydrate (C,H,Na,0,e6H,O) in 2 liters of deionized water. Mix 
the plating solution by stirring until the solids have totally 
dissolved. 


pter the plating solution with a lint free filter. A 
membrane filter (lym pore size) used under vacuum is preferable 
although a glass fiber filter may be used for gravity filtration. 


d. Areas of the sample not to be plated should be covered 
with micro-stop and allowed to dry for 24 hours. Clean the 
sample to be plated with an appropriate solvent (isopropyl 
alcohol) to remove any oils from the surface. Do not handle the 
sample after cleaning. 


" М.М. Morra, J.M. Morra and R.R. Biederman, “A Technique for the Preparation 


of Powders for Examination by Transmission Electron Microscopy", Materials 
Science and Engineering, A124 (1990) pp. 55-64. 


119 





EC2.8.2 Procedure 


CAUTION: The plating solution must be heated to activation 
temperature in a water bath. DO NOT ATTEMPT TO HEAT THE SOLUTION 
ON A HOT PLATE! 


a. Suspend a 2000 milli-liter beaker in a 4000 milli-liter 
beaker. Pour enough water in the 4000 milli-liter beaker to 
immerse most of the 2000 milli-liter beaker. Heat the bath. 


|pmELour the plating solution into the 2000 milli-liter 
beaker allowing room for the sample without over-flowing. 


c. Suspend the sample in the plating solution. For best 
results, do not allow the sample to touch the surface of the 
beaker. DO NOT COVER THE BEAKER DURING PLATING! 


d. Allow the temperature of the plating solution to rise to 
between 85° and 91°C. 


e. The initial pH of the plating solution should be 
approximately 5.5 to 6.0 (test using pH paper). 


Monitor the plating solution carefully for the first half 
hour after plating has begun (the rapid evolution of bubbles from 
the sample surface). If a white precipitate forms in the 
Solution (it will cloud the solution), add lactic acid drop-wise 
and stir until the precipitate disappears (solution clears). 

WHEN ADDING LACTIC ACID, BE SURE THE pH DOES NOT FALL BELOW 5.5! 


g. Depending on the surface of the sample, the rate of 
nickel deposition will be approximately 1 mil per hour. Ifa 
problem initiating plating on the sample develops, it is 
suggested that the sample's oxide layer be removed by placing it 
in a heated solution (approximately 60°C) of 2% Hydrofluoric 
Acid, 2% Nitric Acid in deionized water (by volume) for a few 
moments. The above procedure should then be repeated. 


h. When excessive plating occurs on the glassware, remove 
the sample and cool the solution. Additional plating thickness 
may be obtained by repeating the above procedure as required. 





À-3. Detailed System Drawings 


The following pages present the detailed system design and 
construction drawings used in assembly of the alloy 600 test 
platform. Detailed listings of components are included with a 
listing of the manufacturer from whom the part was procured. 
Alternate sources of components which possess similar 
specifications may be substituted for those listed. Several 
components were custom assembled within the lab or were 
previously assembled and the source may not be known. The 
layout of connecting tubing is provided as a guide. The exact 
dimensions and bends were made to custom fit the valve 


Оса опе on the test stand boards and on the autoclave. 


Post-assembly testing consisted of system hydrostatic 
tests using the installed pumps as a source. A Heise gauge and 
digital pressure meter served as the pressure monitoring 
equipment. A hydrostatic test using distilled water was 
considered successful if no leaks occurred in the system under 
test for a period of thirty minutes. Components which would be 
pressurized at temperatures greater than room temperature were 
hydrostatically tested at room temperature and at their 


operating temperature. 


121 








Jv 19 HO43 NMOHS LON IN31SAS NOILO3ÉNI LON 
—— — — ——JdMfld NOILV'InOWI938 


— —— —MWOL1v1n93M (9I1V1S) N390HOAH 





E 


7 MWOAVTF0938 TOW1NOO WIV H31SO008 SVO 


QHVO8 IOHINOO 
—-W3ISAS OH.VlS ANY NOLLVOIDnIONIO 3 


10.83 


QNVIS 1831 NI 3AV10010v — — 


SNOILO3NNOO ONITOOD — 





QuvO8 1OHLINOO 
-NOtLVZIHRSSdHd N3OOMXLIN 3 dlAVS 


ALlV 19 HO3 NMOHS S3'ddlN 2 ATNO :310N. ——— 





ANVL 3OVHOLS IILSNVI 


(c WNVd AV MSIG залази ТУЛЫ 
Ба 


Alloy 600 Test System 


Figure A-3-1 


122 








Sample Nitrogen Board Showing Valve Layout 


\ Drill 1 hole each side 
at 0.5" from center 


0 075*— 


17.507 








Sample Pressurization Board Drill Template 


Figure A-3-2: Sample Pressurization Control Board Tubing and Template 
Details 


123 





Autoclav 


















о 
E 
=. E = 
ОЕ 
> o2 
same 55 a 
= 
Y Loin 
= Y 
15 5 
> Ow 
(а 5 
n об 
- 5 | E 
mim | ОФ 
CI 4—3 | O > 
|| 7 ow 
IB => 
O о. So 
IE Е o = 
С) о 
RE ==>" | а. с о 
1 Sa | о Gc 
|| = со 
о 5 
| on и © 
Pre Р Т! E = о 
ш, O OOK HE cpa Б ==== о ва 
qc 
802 
9 
c 
Ф 
D о 
Be 
58 





Ф 

< 

S © 

1.50" Ф с. 
& E 

o o 

E m 

E Е 

= а 

à 5 

© о 

Ф | e 

-1.50" о б 
m g 

Е 19.00" w 

2 

Е 2 

© Ф 

3 С 

o 

5 © 

1.50" Ф O 
x 9 
w 

o 





Figure A-3-3: Recirculation and Static System Control Board Tubing and 
Template Details 


124 














Ayddng any 
pesseuduio?) uio44 





1ueA episno 04 


unge Buijoo?) - | 


> Aiddng биңооо L 


о ИЩИ 


Figure A-3-4: Alloy 600 Test Platform Schematic 


123 





Alloy 6UU lest Flatiorm Componen IS 
grass | dTenoid Operated Two Way Valve 
р ше {[ АЯ Engineer 
opper 
Brass Minuteman Controls 
ainless Stee 


ainless Stee Boston Hydraulics 0-10000 psig Liquid Gage 
ainless Stee Boston Hydraulics 8 Street Tee -Hydraulic Fitting 
High Pressure Equipment {Autoclave w/ Cooling Jackets on Standoffs 
McMaster Carr 
P ingle Stage Air. Regulator 
Midd 00 cuft Hydrogen w. tage Reg 
Midd 00 cult Nitrogen lank, Grade 4.8, w/Connector 
rove as Regulator 
00 cuft Grade 5 Hydrogen Tank with Connector 


а 
О 
C 
O 
A O 
e 


ainiess otee 


c" 
ө 

C 

© 


Hydrogen, Grade 
Nitrogen, Grade 4.8 
Brass Case 
tainiess Stee 

ydrogen 


ainless Steel | High Pressure Equipmien 
ainless Steel _| High Pressure Equipment |178" Tee Type Safety Head] 
- D 


= 
О 
О 


Г - 5 
(> 
(> 
(0 
O 
2 


ON © 
бој ОЈ 21 


| 5000р50- | Stainless Stee 
| 10000р50 | Stainless Stee 
| 10000 psig | taınless Stee 


600 tainless Stee 
6600 ainless Stee 


| 


[ S600psip | Parker 

| 10006 psig | ainless Stee Parker Union Cross - 1/4” CPI fittings 

[ 5000psig | Stainless Stee Parker 

[ S000psig | Stainless Stee Parker 

10000 psig | Stainless Steel |  Autoclave Engineers 
l 


900 Brass Parker MPT to 3/8 ex Nipple 
| 20000 psig | arbon steel as Booster Pump 
000 Polyon Aeroquip Polyon High Pressure Gas Line 
2 0 1/2 
PIto 


000 tainless Stee 


200 psig | PalyEtaylens 


500 ainless Stee 


[ 0000 psig | Stainless See 


| 10000 psig (1) | opper ustom ooling Col 
Backpressure Regulator 


— 
(UJ 
L] 
O 


Ц 
С. 
D 
e 
O 
O 


Omega Engineering  [0-15000 psig Transducer 
Omega Engineering  [0-3000psig Transducer 


| 60000 psig | ainless Stee Autociave Engineers 4 High Pressure to 1/8" SpeedBite 
| /900psig | ainless Stee Parker 4 MPI to P| Male Connector 
|. 7200psig | ainless Stee 8 MPI to 1/4 CPI Male Connector 
| 10000 psig | otal AME 


МРТ Н 
Non-Conductive High Pressure Hydraulic Hose 
D 
e 


ainless Stee 


g u S 
O O 
= > ГІ) 
аж 
a 
O) > 
© 

О 70 = 
[1r] 


а 
2 


0000 tainless Stee Boston Hydraulics [3/8"FPTto 14" MPT Bushing < | 
0000 ainless Stee Parker Male Connector - 1/4" MF T to 6 CPVA-Lo 


000 taınless Stee Autoclave Engineers Way 1/8" Isolation Valve 


08 
E | 10000psig | ainless Stee Parker 4 CPI/A-Lok Union Tee 
7 000 ainless Steel - 


ОЈ 
а 
e 


ШЕ 


| 


0 
Autoclave Engineer ее - 
Parker 173" 





Pl Fittings 





Table A-3-1: Alloy 600 Test Platform Components Summary 


126 








65 


64 


17 





From Compressed p ees 
Air Supply 


ja 














Figure A-3-5: Sample Pressurization System Component ID 


127 











35715000 psig | Stainless See Welln 
[5 | 51220 | T1000 [Stainless Stee 












Autociave Engineer 





[Meilen — — 
[Autoclave Engineer — 
[7 | —30Vz081 | 11000 psig| —Staimiess Stes Way 118" Isolation Valve 
[ 10 |  ST2220 | 11000psig | SiamlessSteel | AulochaveEngmeer | Tes V8" — — — — — 
[1 7] — | 75000 psig_[ Stainless Stes 0.125"x 035" Tub 
[—27 | —5M42B! | 50000 psig | Stainless Steel | —Aulociave Engineers —| — —1/4" High Pressure to 1/8" SpeedBite 
[28 | Р605х | 22500psg |  17-4PHSS | OmegaEngineenng | 0-15000 psig Transducer 
29 | Pex | 22500psg | —174PF 0-15000 psig Transducer 
[30 7563AF2 "| 15000рзо | Stainless Sies [High Pressure Equipmen 


[31 | | e000 psio | Stainless Stes Meta Men 0-257 x 035” Tub 
2 | тыш amless 51е Metal Men 0.7257 х 0357 Tub 


OV 000 
KIP 14116 00 C 
000 
000 


Autoclave Engineers Air-To-Close Diaphragm Valve 
Minuteman Controls ree Way Solenoid Valve 


ainless tee 
Brass 

ainless Stee 

tainless Stee 





а 
= 
C 
O 


Autociave Engineers Way 1/8" Isolation Valve 






O 
< 
O 
00 


а РВ 00 ainless Stee Parker 4° MPT to PI Male Connector 
8-6 MHN-B 900 Brass Parker VIPT to 3/8 MPT Hex Nipple 
5 ET | Brass olonoid Operated Iwo Way Valve 
6-6 FB 00 ainless Stee Parker 3° MPT to CPI Male Connector 
48 | |10000р50 | Stainless Stee Metal Men 0:25" x 065" Tub 
| 49 | | 8000 psig | Nitrogen, Grade 4.8 Vida 00 cuft Nitrogen Tank, Grade 4.8, w/Connecior 


[52 | | T0000 psig | Stainless Stee 
| 64 | 000 ainless Stee V V 0. х .0 ubing 

Ecc Му | сеп | Alloy 600 Sample 

| 68 | 2478255 | 7500psig | Stainless Steel | — ^ Parker | 4 MPT to PT Male Connector 
IMSS] Nickel 200 | High Pressure Equipment ^ Autoclave WT Cooling Jackets on Standoffs 





x .0 ubing 
Metal Men Ö 06 b 












Table A-3-2: Sample Pressurization System Component ID 


128 





20 











Figure A-3-6: Static System Component ID 


129 


14 


23 











Р 0000 
0000 
6000 
0000 
0000 
0000 
0-\ 0000 
0000 
9000 
UUUL 
0000 
0000 





Table A-3-3 


[UID| —PartNumber | MAWP | 
O AZABZOZSSP | SU0Upsig | Slaimess SIee 


LLL 
LS [ УБЛ | B000 psio 
[T0000 psig | Stainless Ste T" Male Connector 


m | 


| — — — [095g | Fyaragen —] MidüserGas | зоо арта 





atic system Component Identification Li 

lea a a : 

мо Уау Ва Valve - ittings 

_ ттт [дете — Aeee ae — 
Parker 
A" Male Connector 


























55" | Parker | — —O-RingPsppelCheckVave — — | 


ainfess Stee [Weta nen I ZERO Tan — — — 
en | Farker Ring Poppet Check Valve 
Hig 


ainless it | Boston Hydraulics | Non- ae ive Pressure Hydraulic Hose 
ainless Steel | MietalMien | 0.06 
ainless steel | МеіаМеп | 0 0.06 D 
ainless кюл уле reet Tee -Hydrauiic Fitting 


ainless Steel B 0-10000 psig Liquid Gage 


Hydrogen lank wi onnector 





: Static System Component Identification 


180 





vn Ф 
pon | "Ei BM og 























б--------------- 


0 eni 
cS ча m 
221% | | 
Жос 
19 Log 
ws 
шәл IPISINO OL 
== 
s — 3 
S 
IS ~ 
о == = 


| 


Recirculation System Component ID 


Figure A-3-7 


131 












neciculation system Component Identification Lis 
EA osa | ште | Custom 200 iter Storage Tank w7 Sight Tass 


ainless Stee 4" СРІГА U D 
ainless Stee Diaphragm Recirculation Pump 
ainless Stee 4" Male Connector 
ainless Stee wo Way Ball Valve - 1/4” CPI Fittings 
ainless Stee wo Way Ball Valve - 1/4" CPI Fittings 
0000 ainless Stee 0 


ИИ ЖИ т шора | Stainless Stee 
Ne 444у)52-55 | 1000р | _ Stainless Stee aan 









Pulsa Feeder 
Parker 






с] cu = = I<] <j <] <] <] са <Ј 
т О О J U U 9 
< а са < A SI а exl «| cj «| с.) <, 
- һ] О 
O 
O 
O 
С 


PI/A-Lok Union Tee 
ainless Stee 0 0.06 D 





U 0.00 D 


[20 [10000 psg | Stanes Stee 


[FAT IBZ-SS—[~T0000psig [Stainless Stee 


ШЕ А 00659. | ee 
ПВ ЕЕ SERERE 


PI/A-Lok Union fee 
0.06 


PI/A-Lok Union Tee 


C . 
- 


h 


ainiess otee 


| 32 | 4-4FEZ-55 | 0000 ainless Stee 4 Maie Connector 

Two Way Ball Valve - 1/4" CPI Fittings 
[3710000 psig | Stainless Stee 0.25" 0.065" Tub 

| 38 | 00 000 tainless Stee Backpressure Regulator 


[< | TO BIAFA | 10000р50 | Stainless Steel — | High Pressure Equipmen i" Tee Type Safely Head 


[41 [| 10000 psig | Stainless Stee : 
i8 | WA [T0000 psig 7] opper 


0.06 
wo Way Ball Valve - 1/4” CPI Fittings 
o 1/4 CPI/A-Lok Male Connector 






ud 
J 


Parker 


D 
ooling Coi 
High Pressure Equipmen Autoclave w/ Cooling Jackets on Standoffs 
Alloy 600 Sample 
Parker Union Cross - 1/4" CPI fittings 
4 CPI Male Connector 
0.06 
4" OD P 
4 CPI Male Connector 


A 
O 
C 
c 
= 


p 
O 

O 
O 


00 

= 
U 
o 


ainiess oiee 





[| 10000 psig | Stainless Ste 
| [| 200 psig [Polyethylene 
[ 55 | 4578 [| _ 7200 psig | Stainless Stee 

ainiess Stee 4 Tube CPI General Purpose Needle Valve 
[11] 10000 psig | _ Stainless Steel | MetalMen | 025 х0.055 Тиртд | 


O 
= 
O 
O 
00 
= 
О 
e 
o 
- 


= 
> 
С. 
2 
C 
> 
C 

c а 

О O TU О 

= 54 

C 

Ф 

O 









Table A-3-4 : Recirculation System Component ID 


1592 














Figure A-3-8: Injection System Component ID 


133 


























njection system Component identification Lis 
5] 3008030000509 —| чате яве | DC Anaya — [Min UEG, Pump 

ops Е MS en Tn — — — 
— 2000 psig |Hysogen Grade — Medleser u — 00 cuf Hydrogen w/ 2 Stage Reg 

Бє с: жк от: Бы | а a AR AA Ни 
орлы Tcu — Worms И 
25 OOPS STC] Моего | High Pressure Equipment |Autocalve w/ Cooling Jackets on Standorts 








Table A-3-5 : Injection System Component ID 


134 








& 12 
11 
2— 
13 


MH 


Figure A-3-9: Emergency Depressurization and Safety Collection System 
Component ID 


> 












calado оп апа satety Collection oystem Component Identification LIS 


AL ASFRESS | 7200507 amiess Steel | Parker Vale Connector 
ны | 0000 ainiess olee Vietal Men х 0.06 чота 
| 3 | 4-4-4 MBZ-95 | 0000 ainless Stee Parker emale Run lee 
[8| Copper [Custom [Single Element Cooling Coi 


0000 ainless Steel | Metal Men | 0.25 x 0.00 ubing 
[TT | 4-1 WBZ-SS | 10000 ainless Steel [aa Parker 4 Bulkhead U 











Table A-3-6: Depressurization and Safety Collection System Component ID 


136 





АНЯУ1Э HO3 Q3NOILO3S LON SEN LHƏIY ‘ALON 


MIA 1NOYW ,V - V NOLLO3S MIA Adis 
real 


Д 

















m їй Е 
PD fi D m 
| | ES ma ds 
[E — ЕАМ 141009 e О) „бич до „ег 
E 5 m г. Ww t 1 ә20029 „S2'0 
= | EN Bag 
т -L340 1NV1009 O). (О 
mes ІШ ШЕ 


м ozeJg pue o|ddIN јој 
Р 9jou ,£9'0 uBnouu] Iud 
—— део aqn| Jaddo) ,G2'0 


IT | 
Qmm | 
шы 
ПОЛАЧЕ 


„егу 


5Ппауўуч „56 1v dados Sa 10H 8 





2 8.0000"—— 


-3-10: Autoclave Head Detail 


Figure A 


EST 





AULNUS 1284 341 џезш рџелџр 92 е чим 


d 


реән елејзоулу реән әлејојпүу —— 


uBnoJuui uoneljeysu| eqn | eiduies Jo sııeJoq 
BuIMOyS y91 ој еолу раролу) јо ме решибеуј 





«96 0 == 


г |еощэәјә 104 
әәр шо | рәѕц иоцәј ХУМОЭ шеан — ^ 





би ,SZL'O XVNOO — 


џуеозимџоје | ХУМОЗ — — 


A 


san+ ојашес SZLA 





эмел зиэл —— | 





Autoclave Head and Sample Assembly Detail 


Figure A-3-11 


1339 





4C€0 0 0) peuuiu| 





8-g uon»es deg woog »2612-eJd әпбце4 JOY ПеЛЛ Jeuu| uM 
ојашеб 159 jo uorjoes ојашес 159] 10 иоцоес 
E | 
i I | 
V-V uonoes ' 
In | zi nr 
aıdwes 1S9] 009 Коју Jeoeds JEUISJU] DILUCISD 
= 
S 
| 
P3JON 9si"ueuiQ SSAJUN 
009 Áolly ese sjeuajey ajdwes jy — | 
zr E 


umoys 56 
'so|d £ PIM 





yyBue| „81 о] d 


Buiqgn] [9915 SSaJUIeIS ,SEOX SZLO' 


Test Sample Assembly Detail 
139 


Figure A-3-12: 





SUILFAN ЭІӮ-аа VOINO 

























SGIION3 OS 
[ алпа 4315009 IIASVH ANY 
Z-dOV Е 9d | L-dN | p ay | сен JAVA dANA NIDOYLIN 
A у I 
[| 9m - — | ТІРЕСЕ 


HOLIMS | 
3ülH83AO W31SAS | 


z1sos (1505 


гато 
bd 
1-2 
vio 
Pr st i or 


U3MOd YZTIOULNOI 
39485 ИОН — H31V3H OL 
NI AOSZ 1NO A0SZ 




















eseeees 





г 




















5) 


| onmin 





| HOLMS 
9NILVH3dO чипа 






HOLOV.LINOO HO1V1 


Figure A-3-13: Detailed Control Wiring Schematic 


140 









AMO y 600 Autoclave System Control System Wiring 








VA ervice Power to Power 


rip 
onitro! Lead to »ystem Override омс 


VACROP ТИ р ее вова — — | 
BRC SS VAC Service Power to Solenoid S-T(Haskell Pump АТ Shuto 
AA Neutral Lead lo Solenoid S-T(Haskell Pump Air Shutof 
[8551 | тыз | — 45 —— round Tor Solenoid S-1 (Haskell Pump Air Shuof 

VAC-PLU | АНИ 

AA 
ше: y 
ISC | 
[50572 | 
ee 
e 


p 
U 






NP- 1(P6)-N ontrol Lead to P6-NC on Meter NP-1 via Jack J-6 
C-LP-3 ACP-2(P6)-NC NP-1(P6)-COM Series Jumper from NC Terminal on Meter ACP-2 to COM on Meter NP-1 
[ —TBZ16 | ACT-APS-COM | ontrol Lead to P6-COM on Meter ACT-2 via Jack J- 


C-T-3 ACP-1(P6)-COM ACT-1(P6)-NC Series Jumper from COM Terminal on Meter ACP-1 to NC on Meter ACT-1 
B 9 | АСР-ЦРО-МО | ontrol Lead to P6-NC on Meter ACP-1 via Jack J-6 


Table A-3-7: Control System Wiring Detail 


c 
ci O U 
“ОО 
U U = 






141 





A-4. Axial Fatigue Pre-crack Sample Preparation Procedure 


Initiation times for SCC were found to be considerably 
shorter if the localized stress were increased. Axial fatigue 
cracks initiated in the test area provide the required 
localized stress and a crevice from which SCC can occur and 
grow. In order to evaluate fatigue growth in real time, the 
fatigue crack must initiate at 90 degrees to the applied load. 
However, calculations using MATHCAD®‘s Roark’s Formulas for 
Stress and Strain on a tube sample show stress levels are a 
factor of 1.7 times greater on the inside wall immediately 
under the point of load application than at 90 degrees to the 
point of application. Several early tests demonstrated this 
when axial fatigue cracks occurred and grew through-wall 
directly under the line of load application. 

The stress at 90 degrees to the applied load was increased 
beyond that directly under the load application point by 
machining a thin notch 90 degrees to the load line with a 
ДОНЕЛО tool and a 0.875 inch by 0.015 ineh cutoff wheel. The 
depth of the notch can be estimated using figure A-4-1. 
Typical depths to initiate fatigue cracks at 90 degrees to the 
load line vice under the load line ranged from .030” to 0.035” 
in the Low Temperature Annealed Alloy 600 Heat 96834. Fatigue 
crack growth is induced by applying 90 percent of the yield 
stress load (approximately 1000 pounds) at 90 degrees to the 
machined notch in an MTS Model 810 Material Test System. To 
ensure even loading, the sample is held in a fixture 
illustrated in figure A-4-2. The MTS Test System is operated 
in a sinusoidal mode with minimum loading not less than 50 
pounds compressive to prevent sample movement. Figures A-4-3 
wd A4 4 illüstrate the actual test setup. Initiation of 


fatigue cracking is noted at the edges of the machined notch 


142 





using a stroboscope and microscope to monitor the notch area. 
Typical growth should not exceed 0.1 mm to prevent excessive 
Aires wall crack growth and typically occurs in under 10° 


cycles. 


Chord Length vs Penetration 























Chord Length at Surface (mm) 















































S 55 6 6.5 7 75 8 8.5 9 95 10 10.5 11 
Radius of Cutting Wheel (mm) 


— b=.254 mm 
77 b=.380 mm 
— 6=.500 тт 
—  b=.635 mm 
C7 b-2.762 mm 
77 b=.889 mm 
— b=1.0 mm 


Figure A-4-1: Notch Depth versus Surface Chord Length 


143 





^ TUBE SAMPLE 





G-10 BLOCK MATERIAL 





Figure A-4-2: Schematic of Fixture Used With the MTS Testing 
System to Induce Fatigue Cracks in the Test Samples 


Microscope arid SOLET ON Y 
SgpobcAsemb ; а. № "Sample Jit 
"m | Pikture toad 
Sample Mng a 
Single БҮ Ji 
Line 


+ A 





Figure A-4-3: Side Photo of Sample and Fixture Mounted in the 
MTS Test System 


144 





i 


y 
Sample b 
Fixture - Front 
View - Installed 
in MTS Test. 
System 


Figure A-4-4: Front View of Sample and Fixture Installed in MTS 


Test System 


145 





A-5. Error Analysis 


An error analysis was performed on the crack growth rate 
and stress intensity data obtained from the tube tests. The 
errors associated with crack growth rates and stress intensity 
are illustrated in Section 6 (Discussion) and presented in 


Table À-5-1 belov. 


Errors associated with the stress intensity calculations 
consisted of analytical errors from the finite element 
analysis used to calculate the stress intensity. The 
estimated error associated with this analysis is +5% The 
plotted error bars in Figures 6-3 and 6-4 represent the range 
of stress intensities throughout the crack growth period with 
the upper and lower limits accounting for the 5% analytical 
error. Crack growth from a free surface is assumed to be 


dependent on K, when stress intensities reach a value of 4 


MPaWm” and have no associated analytical error. 


The error in the crack growth rate is primarily a 
function of determining the initiation time from the ACPD 
output. The determination of the initiation times is 
discussed in detail in section 5 (Results). However, in the 
case of surface crack initiation, the sensitivity of the ACPD 
may preclude detecting initiation until a large enough crack 
has propagated into the vall. Therefore, the worse case is 


initiation from time 0. The lower values of crack growth for 


t! J. C. Newman Jr., and I.S. Raju, “An Empirical Stress Intensity Factor 


Equation for the Surface Crack,” Engineering Fracture Mechanics, Vol. 15, 
Eo 2 (Creat Britain: 1901), ор. 185-192. 

? Electric Power Research Institute, Steam generator Reference Handbook, 
EPRI TR-103824, Project 2895;4044 Nuclear Power Group (Palo Alto: EPRI, 
Dec 1994), р. 13-26. 


146 





the first series of data reflects this estimate. However, the 
rise in potential on the 10kHz signal is most likely the upper 
limit and this is the plotted value, so there is no associated 


upper error. 


In the case of the pre-fatigued samples with multiple 
growth areas, the average growth rate was plotted from visual 
measurements of the upper and lower bounds of each region. 


These upper and lower bounds are then used to determine the 


upper and lower growth rate errors. 








Alloy 600 Test Data with Error Limits 


Test K (lower) | K (upper K (avg -) Error +) Error | (Average lower upper -) Error +) Error 


EUER 73 | с; | 17 | 20 | 46-3 "41 | 112 | оь | о Ц 
КОЕ | 92 | ee | 268 | "91 |_> | 13 | 436 | 07 |_ о 
Em. "| вз | 62 | 22 | 26 | 24 | 44 | А | о7 | 0 
[950909с a | 91 | 66 | 26 | 30 | 23 | 15 | 156 | оз | о | 
ШЕШІЗДЕ-4- | 453 | 97 | 57 | ба | 27 | 18 | 85 | 05 | о _— 
CEC EVA 09 | io —|- 36- |- 18 | 9 | 5 |_ 23 — 











Table A-5-1 : Test Data with Error Calculations 


147 


‚- КӘ 


БІМР5 
an 
/99 22521-2 


c2 


me 
У 





ii 








on. A А) 

t reste Аы 
2 

Ў, 


n 


S 


МАЛАРЫ АА 
Eier 
ERR ario ae 
EN 

М Д 


tI 
NUR) 
s 


P 
En НН A 


ee 
м, 


Бора iite 


ИРА А rus 

а pens 

de c 2), ДЕДА 
IL "D 

аа a 


3 


i 


ЈЕ 


5, 
Жі ун 


LIBRARY 


ІШІШІШІ! 


54766 2 


AC E 
С, 
ИРАНА 


n 
Ut 
ih; 


acs 3% 
ELA 
rr 


Pe iot A 


A 


af Рао un К 


који 
ЫЛ 


iw 
Кони 
X d^. e PIS 


van 
T E po: 


A e 


ASIA 
ES 


Ё 


D 


ARM: 


р tna dal vn Uy han 

"A 

Ё MEINTE 

АН 
oe PE TEM 


E к 
АИ ^ Ц 


Dru pen vette TP Ж дігі 


DEAD 
ү А, УР 
] 


CR RA e 
Werth akon etd 


E ie 


Vy 


V A 


AS 


oL OR UON 


A 
ELS 
VALE 
Nr 


C 


з 
ја 


9 %ы 
FA E sa 


Er 


yy 
en 
А Pa 


d i 
нн 
n 


n 
S 
ЖЕП 


wur 
m 


TORET У 
ығ”. Ar 


ЕТТІ 


it 
DH 
ОТЕ 


э зт та 
ПЕЦ 


$, 
E 
ТЕЛУ 


E 


E mn 
ж у ШАҢ Mn A NE 
(teres ter А 151 
АКУ 
A 

un" 


B 
" 


НА TS 
А АИ 
ПИ МАТА Еи 


H 


е 
2545, 


ГЕ 
ботев 


ауы 
lt sa i Ya 
4) 1 vee 4 у HET 
SAM HON қ Mon vw , p JUL ar 
Bre d TR ТУТЫУ 
" 14. 3 


ye Ы 
M LA 
ғ 

yo HE lem Жа 


un р 
b nen ni 
NN. 
Н 
СЫ? 
ІТ 


ҒА! n ы 
Til 
ans 


D 
Н КЕЛІНІ H 
Pi ni 3, 


Hi 
Ws 
"ur 
ГА им 
ү 


р | ОН 
НЫ mal | «BT 
HEN PER i i КИНЕ, НУЫН РЈ 
ан АН PEIPER M ни us A 5 uM fi 
e 1 Urn 


DP 


i 
у 
XH 


f 


А | Џи M Ú "n, | «и yen ees 
XU. К М и \ AUR) 


ООУ 


14444 
Й Ж] 


р 
| 
Й 


ИСТИИ 
oh УК ГЕ 
., П О 


ЛАО 
бі. 
t Й 


онага 
UAE SPD