ML20141K408
Text
{{#Wiki_filter:* ( l February 16, 1968 shs.AA, N.e Sif MEMORMEDOM To : AC'RS Messbers From t Barold Ethertagton, ACES Member subject: OYSTER CREEK FRESSURE YESSEL SWI TUBE TM1ERES nts memorandum is presseted by the consideration that repairs to the Oyster Creek vessel will be essentially complete before the Casemittee has had an opportunity to study and appraise the stress analysis by the contractor and his subcontractors. I teason for Caecern he basis for concern is that the contractor does not appar to be con-carned that the stresses are autremmely high. Contractor's Statement on Cause of Failure. ha contractor (Raferente 1)
. steributes the problem to " chemical activetion of the surface of sensitised i
stainless-steel stub-tube material contatuod senerally within high-stress areas of the stub tabe, the presence of defect-sontatatas field welds be-
, tween the stub tubos and soutrol rod housings, and minor defecta contained i within the welds between several of the in-core tastrument tubes and the ,
vessel". B ere can be no quarrel with this comelusion as a statement of l probable fact, but the statement fails to indicate that the strass is b extressaly and perhaps umaeceptable hiph. I contractor's statement of Cause of stress. Reference 1 (Ammad= mat 19) met only touches on stress quite lighcly, but it attr N tes the dama41ag stress j to the shop weld. On page 18, the a==aa-ant refers to 'All of the cracking indications associated with the shop weldius ..."- this assentially ===ina all of the erecks at the bottom of the stub tubes. I halieve that, although the stresses from the shop operations are kish, they are overshadowed by stroassa from field welding. It was for this reason that ameasureensts of tube diamatars above end below the weld were requested and have been emppitad by the contractor. OFTICE > dCN ......-. . . - - ~ ~ - - - - - - - - - - - - * ~~' " " ~ " - " " - " ' SURNAME > . . - - - - + ~ ~ ~ - - - - - - - - - - - - - - - - - * --
~ ~
DATE> ----------.r -- - - - - - ' - ~ ~ " " " " ^ ' " ' ~ Fbrtn AF.C.8tt her. 9-43) p.s gove,urm mwTms ma tm.-o ai4 4m 9203190195 910807 ,3t ' PDR FOIA - Q l ') L DEKOK91-282 PDR a
( ( l i swever, neference 1 (the February letter) statu os p.3 ht "saporteses and mensuronete ef tha distortles indisata thet b fLold weld teseen b stub tube end the eontrel red drive houaing syy1ied faketeatice indueed settas whish emeeed the elastic ilmit of b material en the ' free' lassth of $a ,
' kill eide' en & outer row of tb stub tihes". (he ondereescing is adas,)
Wia tediaates a shif t in & sontraster's posittee, but he has met resspateti (or has not acknowledged) Wt th salestated stresas sa tba laag side, s1 though smallar than those en h short side, also encoed the " elastic 16mst* by a wide isargin. N proposed removal of weld metal (pp. 3 and 4) will beve a beneficial effect ou strates, partly (as stated) beemsee it wit' imereene W ef foctive langth of N stub tuba, bet also (not elated) kasuse it will weehan the weld area euf ficioatly to permit yieldtag and relief of boeding savent-this seay be a dubious remedy. i %s contractor now proposes (Reference 1) to resore M field welds. H is is fine, but there will still be high residust stresses although of eyposite sign. It is hoped that, in developing procedures for rowelding (pese 40 of Esference i 2), the laboratory will pay attention to avoidance of the sucessive deformation evident from field measurements -- this is set mentiommi in the reference. It is concluded by CE*,-CE and Telodyne (Reference 1, p.3) the "the stresus
, tend to be congressive in nature in the hot operating condition". It is true l that there would be substantial relief of stress from the shop weld at opers-ting pressure, but again the et.atement appears to ignore the ef fect of th
! told weld.
. Conditions for Stress Corrosion his is a problers for the metallurgists. However, it is moted that on p.3 ef Reference 2,it is found that 10 ppe of chlorides produced crack.ng of sensi-
!' ~
tised Type 304 stainless steel " stressed above yield" in 48 krs at 1800 y,
! h is raises a questiont Row much chloride would be required to have a similar i e f f ec t in 350,000 hrs. at 5500 F, and possibly et some higher stress and with other unpredictable variattoes? Will the proposed eladding of Type SOS L pre- -
i vide the desired protection and imanuaity? l j _ Int erpretation of calculatM strass l Design calculations are based on elastic theory. In asses such as those weder coc. sideration, the calcuisted stress le fictitiously high, because plastic defersu. tion limits the stress to a va'.; somewhere above the yield stress but i usually noch below the ultimate strength. All stresses given in this memorandum are calculated from elastic theory. Calculation of true stresses would require a vary complax elastie-plastic enslysis. As a more realistic first approximatica to the true stress, the calculated stress saa be converted to strain by dividtag by the modulus of elasticity. 18 a 136 pois for ==ple, a salculated stress of 3$0.000 poi corresponds se - a strata of 0.0125 in./in. In a plain carbon steel, the actual stress would he If at tad by 31.ne te defar=neir% *a pravNa ply the y41@iat,--4m = austpattle steel the actual streasjwould be coniiderably graitar than the yf0/ *s Na a , "and ' could be~ da tesid:sd~ froe' tkGtiks e-s train curve , su m m > _ - -
,.)
== = = = = = = = =
M n.AECR M n && g m ,,c.p,w,, e , m .g3
l L ( i l
. 1
-1 ne strus-strata relation for as susteattle steel is aan linear spee belas i the yield stress. 1.f tbc 0.1 percent offset yield strength der AISI Type 306 stataless steel is 35,000 pai, them the total strata, elastia pins imetas0ts, _ l st the ytald stress is 0.00325 in./in. Bis means that a salestated stress . ,
of 91,000 psi would La focc (to a first opproatmetton) eorraspeed to en aseed l stress of only 35,000 pai. This steel, however, has no sharp yield pelate and strusu such as those calculated in this memorandom will lead to estaal stresses ameh above the yield strus, it is established that .streertion is the stress sriterios in strus seerosion, or is strain also a factor?
' Deaian Criteris j $<mme clarification of the design criteria might be in order. page M of Reference 2- (Amendment 29) states the vessel design "complias with Section 1
. . . of the code, with applicable Nuclear case Interpretations, .. . . In addi-tion, the specification requires a detail strcas analysis cooperable to that i required by the-1963 Edition of Section III Duclear Yessels but with the material allowable stresses as set forth ta section I. .... The General Elaet.ric speciftestion requires a fatigue smalysis ... "
I Allowable Stress, ht is the maxfanum allowable local stress or strain
- Dnless
! it is proposed to remove all esiling on developed strus or strain, abort of tensile failure, some criterion would appear appropriate. In which eategory under Section III are the bending stresses treated.
Code Interpretation. For obvious contractual reasons, the Code sousiders a vessel to end at the possles and does not seasider the effect of field welds. i In this case the stub tubes and control' rod drive housings form part of the pressure boundary of the vessel and, if the housings had been attached in the shop, the detail would have been snelysed and found unsatisfactory in de-ign or welding proeedure - at lesst the fins 1 aasaudily would hers been atreae l relieved. It might he asked whether the intent of the Code has been satisfied j on whether a loophole has been used by which the contractor has. found it com-
- venient to complete the pressure boundary in the field and has not applied the Code - obviously there was no Latantion to evade the Code, but it does appear l,
that an unsatisf actory condition was act recognised by the designer or by the I writar of welding proceduras. Fatigue Analysis, ht residual stresses will be used in the f atigue analysis' y
)
I
.l . .
. ,. . ~ . . .
_ - - . . - . . .~ omet > . . . . _ .- _ l I .- .. . . . . -
$URPt AME > .,
D ATE > ..m.._...a .
- m. . . . . . . . . _ . _ _ _ = -
Form AEC-814 (ker S-53' u s boam M Parms oren ise-<>U. es
( ( Attaehments Attselmusr,ts A, B, and C give the gist of the salculated strass patterms espeeted in the esotrol rod drive housinss and atub tubes. The calculattooe are isome6 on elastia theory with simple models and on the field measurements of the W Creek and 3tagare hohawk vessels. The studies do not purport to be a ser(see stress smalysis of the corditions, but rather a random probias into the seeses and distribution of stress; with the object of raising some guestions whiah may nead answers. The methods of calculation are indicated eely broadly; and da-tails are essitted to avoid further 1 screase in the bulk of this alr**dy leasthy memorandum. Attachment A - Caneral Formula for 1inz loadas Tubea and teferonee Dimmastw. This attachment gives the Sentral formulas for a tube under radially symmetrical toed ing . (A brief explanation of the basis of these formtlas was supplied to
, one A CKS member a t hi s reque s t . The explanation could be written wp if others are interested.)
ne attachment also gives reference dimensions from page 24 of Reference 2, I and the derived section constants. Attach:nent B - Sf esses Induced by Field Weld. % is attachment correlates
! stress calculations with field measureu. ants. Calculated curves of distortion i
confor:a in shape to the measured distortion profile. The peak calculated stress
; in the control rod drive housings is 380,000 psi.
The paak calculated stress for a long stub tube is 388,000 psi. 7or short stub tubes it is many times greater, but the simple calculation 41 model is poor for such tubas. The maximum stress in the stub *.ubes, tensile at the outer surface, l l is approximately 1 in, froes the fis1d weld, or at the shop weld if the length
! is less than 1 in.
t j
' It is confirmed that the stress frors the field weld is suf ficient to cause cor-l rosion cracking at the observed location, and that this is true trrespective l of stress fece the shop weld.
l Lec_achment C - Stress in Stub Tube frou Ebop Treatment. This attachment shows ~ that calculated stresses frca the shop weld, although high, are smaller than
! those frca the field weld. The strenses arise frees dif ferential contraction in cooling frca the stress-annealing temperature. ,
l
.r o m c4 > .. ___
5URNAM[p i h Ec.33 p[ h~
l I I ( Conetusicas
- 1. We control rc,4 drive housing and stub-tube s,ssembliss es destamed and fabriaated are is a state of excepticaally high stress.
y/fi
',%>yl
- 1. We maximum taasile stress at the surfar,s of the stub tukas is at the skey weld where erschina is observed. 'd' 1
- 3. The stress fra the field weld is greater than that from the shop l
- 4. We calculated stresses are fictit f ossly high and high local strasses are conson in structures in the swighborhood of twelds that have not been stress relieved. But, in a location where in-service inspection will be extremely difficult and repair even more so, a very high standard of cpality assurance j is necessary and the pre-axisting stress condition appears to be unsatisfactory.
I
- 5. Very hip,h stressas exist in the control red drive housing ar; well as
! in the stub tubes. A failure of a housing could land to partial control rod ejection.
I
; 6. The applicant should provide a summary stress analysis sod make a
, frank staterment of the magnitude of stresses af ter the initial field f abrias-tion, residual otrosses af ter reewest of the field wald, and fint.1 stresses 4 af ter the repair. Ho should describe his procedures for limiting distortion during the repair veld.
l 7. The contractor should state what design and fabrication changes he plans to mako in future ranctors. (
- 8. It is suggested that the Cournittee, with the advice of consultants if necessary, should as soon as possible rasch its own conclusions coecerning the significance of strasses in ascean of the yield strass with respect to c l stress corrosion and fatigue.
- 9. The pessibly poor metallurgical condition of the stub tubes (sansittaa-tion during atress a.nealing and welding) <tnd probable adverse anvirorsmental conditions are recognized as essantial factors in t.be observed f ailurse but l are not the subject of this ma:norandun.
i Rafareness
- 1. Letter of Februtry 2,1968, R. A. Huggins of Cencaral E.\ectric Company to G. H. Ritter of Jersey Central Power and Light coupany.
- 2. iW~nt 29 "0yster Creek Welear Power plant No.1, Status Report cm Rosetse Yessel Repair Program, December 4,1967. ,,
ev OFTICE > . . . . su-r > . _ _- . . .. _7 . . . . . . _ ..
-. =-.: =. :.::.: : : 2 = x. =--- :-
_.==:,- : =::.= : ==== =:.=. .:-- : : :. ::zu = = :.v.-~- - U 5 40Vinnyrgt eswirm4 CFriCf N-0 214 419
-- . y. .y.m,c p y ;g
,d O . I 34 ;.g g
- g ;p.
-( h O .
.~. %. .+,,eg..
1
'? A k8'
.e. ( fewe e< [re ,41 /?#19 D N [u ? fo< h<.:i2.: h,. .n:(
. $<' <.c-. _.. . . _ . . . .
. lk , l W f h %
nw/ ['.c401er .l)2%rw ma i] __
- e. W.
t ,,
/ ,. 9
. o
_ ,/,_ _ _ _ _ _ _
! -s ,e- -, e a j f _ ,. %
". . ,,f
- -- 8 l f == 4 f A *s ,
l x , g[' .,.. o M s*
- . g / q __ _
. . . g . . . - /
_ ._ k/ ,_,,,e , i t l~ *^* C s f 05 ;'.':. Av.! c :' .~ex wsf C s<wf,.c s a 7 . . , , . 4. . . ., e. . . . i
~ .,
a d .
~ ~2 ,
e'> p
- 5 f rv - s,c i
*(
't'!~ /
, is f .o e
R.n &w , f V '
.* s s *;
1 ~~ '{i % -d'x*.: 'n,o*4 r Q s
^ ~ " " " " " 'M 4., ,, .. .s S r. M: ..
/ 'b 5 . w ,
~]U .~ / w. ^ ?I
/ - :- C d c.:. ::. .? E,1. . t$ r,. l 0 '0 M *((efj3y.r-y,,.r:
.V,' <
Q.g :,Q<!}l.?it)c.., h ,(.s. )gc.,:} n'l'fM) ?i-{c1p),n n.Q . Pf yunin. c,,; ,, J'i
- =.&-2gTcI(3.o,ax/.:L.,m)ie
'l ' '
1 ty c f.'; . p P(o .&.7.p,ga .y fp c)q. ,wy, : u33 -
., g _ . ,
4 -/. / / - A / b,3 ( *.',8 ' 7.~ ..-_.. . , -
.w!.jt.o p 3
=- *
\/ p q.L' .
y 4, y ~;. ~, H h* c. e{ b 3 djv ,
}, --
*y (
a 3 u ~
~j'p' ~
E l:~ t2 (I 4 9 k '_
~4 Q... .
A; *(<' f%f( Gw -wf ,* . b i. I (">=. ,f. y-Da O it's L,.ed/.c A ~C ' % 'G!ujLu , > < na c 7< n. , ,ur6in /~.~ %, R t -/< ( F.]) ll 2. 37 y 2,/yj O.6125- z.5ys xto# I.001 232,r 5 t b $( !s $o 3'. l$ D 3 0&& 3.3}$~ f). 'fFO f.$!"X10 , ,0. 80l
- g
/
h
?
1 l l
(. i Attachment 3 - 8 tresses Induaad by Field Welds his struht discussee implications of the recently measured distortsma ef eenneel red drtva horseingn, cr.DH,(i.e4, the immer through tubea) sad 806 "
- hVf, latroductjg , yy he purpose is to calculate stresses by elastic theory soeutstsat with the sunamurements of deformation.
Ref erenc e s . W e following refersocas were wood: l
- 1. Ammuheent 29, " oyster Creek Duelear Power Plant 90. 1, Status tapert on Reactor Vessel Repair Prof,ven". December 4,1967.
- 2. "Maseure.nents of Fcur Control Rod Drive Rossings at Field Weld",
! CE memo W. A. Kruto to H. C. Msttason, January 25, 1968. (These snessurements j are for the liiagara Hohawk misel).
- 3. Wree-sheet report of Oyster Creek venel measurements dated ,
February 7, 1963. (References 2 and 3 were supplied with copies of a mesmoranstas dated Jesary 17, 1968 R. L. Tedesco through R. 5. BoyM. The Messurements. De requested measuremments were the I.D. of the CR%, tahan at 1/2 in, intervals, above and below the weldt and in two orthosomal diree-tions, corrseponding to maximu:n and minimum veristions. Similar data were requested for the stub tubes if possible. We large amount of dat.a suppited by the contractor covers the requested rage very well, but no set of measurements covers all the data for any particular ecxabination of housing and stub tube. We analysis is therefore pressated j piece.neal, but proba'J1y nothing is lost by this treatment.
! condition of Assemblies. Nothing is known of the history oc general sendition -
of the assemblies. Most have probably been straightened af ter weldtag sad many have baan ground to remove cracks. It is assumed that the C5tDI diensters were i reasonably uniform before welding, but this may not be true of the stub tubes, which would be distorted by the abop weld to the vassel and subsegr.aat strsas relief. l- _. 4 wnct > _ _ . - . . . ... .. . _. . . . L SURNAME > . _. , ,
> . . = = - = = :::. . _ :: =.= u_-- - .g.=.....
t ( control Rod Drive Itopplags The desired analysic requiras that assaseremosta he smede of maulnan med a6mesess distortion in order to permit assassemat of the af fists of both radially M r sy mestrLaal forces and the flattesing forces aristag from aspasasetry of ses>*0 cube length. Only the four swssurenaients of the 5tagara Mekawk vessel gaudt adegnata esta and those are used in the follovias analyets. @aN padially sytsastric Btstartion. It can h shown that the effect of bending
-nt axarted by the weld is aan11 La eamparison with the af fect of dLees?.
radial force. We radial contraetton is calcuir.ted as a fweetion of distassa from the weld sad the results are plotted in Figure 1. 'tha figure shoems that the measured avert.gs t*adial deflectum prrofiles for the Niagaga-Wahawk haastage conform to the shaps ef th profile calculated by alastic thee: y (weve length
, 6.14 in.). The maxieu:s redist contraction (0.020 in.) in the calculated swrve j was selected ergirically to match t.be measured smaximum for cede's 1347, tir2-35, l
and 30-0). The measurements for CEDH 38-11 show sommeabat less distortion but the shape of the profile is the smae. Tha cyemetry af measured dats shows med { balcar the wid confirm the obser ce of any 3 argr. bendtng mannont induced by the weld. Figure 2 shows the pe3k stress, calnlitad by elastic theory for a medales of elasticity of 28 x 106 psi, to ha 380,000 pai. The fLgure shows the rapid I decay of stress typical of this tyve of structure. We radtal force at the weld for a radis1 contraction of 0.07 in. is 95,400 lo. per inc of circumference, calculated at mean radius. Lika the calculated stress, this force is fictitiously higb.
; Figure 3 shove the data for the twelve outer-etrete assembites reportend for the Oyster Croak vessel. The data appear to show a greater scatter than tho9e for Miagara Mohawk and a tendency to more abrupt decay with distance from the wid.
Factors that could lead to the condition include possibly poorer quality con-i trol of field welding procedures, :an out-of-rc= M condition (maasurements l were cade on a single diameter), or possibly inferior quality of the data. Dat-of-round Distortion. The Niagara-Mohawk data provide measurennents in the
~
l II and TY directions for the f our asserablies. The maximum evality is as follows: I (EDH identifiestion number 07-30 02-35 38-07 25-11 i Maximm difference in dianater, in. 0.032 0.013 0.006 0.005 The directions of measurement in no case corrospond to a radial-tangential criaatation or otherviss suggest that they represent smaximuss and minimum e. ear stars, but there does appaar to be a tendency to alongatico in the radial diraattaa of the vessel, and contraction in the targential direction, as a rasmit et the flattaning. . ~;' omcl > .. . - . . . .. . . l .. w-, . 3 . . ... . _ . .
"~'
" ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ' ' ~ ~ ~ " ~ " ~ " " ~ ~
vom rc..f L7"
__ _._ . _ _ ~_ . . . _ . _ ..._ ... _ _ . _ _ _._ ( l i e i Caleulatloss (not attached) shar thet the ferse required te prsomme the ' measured ute4 vedist iorce deflection end that is thean aerese order of in etill magnitudesmaller. De less than the total distrh.h.. dLfforemmes ! in diameter for housing 30-07, which skewed the greatest est-of-round, sem N pletted in Figure 4. A '
* . y> Q ^ . .
. S tub 'Fube se 3 tub-tube- data are sys11able onty ior the Wiasars-Mohawk vesasi. Seoe data i include measurements of tubes well within the array of tubes as usil es. tubes l near the periphery. W asyummetry of stub-tube length-insroases teamede the l periphery. The calysis is made for varicus assumed lengths of stub tube,
! but the tubes are: assumed to he uniform in temsth. A better model sowid ta-j volve a great deal.more work.
_ Bending. The continuous curve in Figure 5 shows the satculated radial son-traction us a function of distance for a tube of Lafiatte temath, and part of a similar curve for a tube 1.95 in. Long, both for a nazisman radial dis-l placement of 0.08 in, at the field weld. Displacement surves have slee been calculated but not plotted for lengths of 2.92 in., 3.89 ia., eri 4.46 la. In all cases, as suggested by Figure 5, the curves its elose to the curve fcr infinite kg;th, eacept dat departure accurs as necessary to meet the condition of sero displacement at- the terminal point norresponding to the fixed (shop weld) etad of the stub tube. In sumanary, for a given displasement at the weld, tha calculated deflection curve is only sitahtly dependent en length for stub tubes two inches long or more. For still shortar tubes a more radical departure is necessary to meet the end som$ition. As- reported for the CRDB Niagara-Mohawk data, the stub tube smaasurements in Figure 5 show a wide scatter, but are not inconsistent with the calculated conclusions. Hecsurements for tubes well within the array are plotted with crosses, and, as would be predicted frru the smalysia, these fall within the pattern ot data for tubes near the periphary. _l
+
The calculated radisl inward force on the stub tube end the maximuss bending stress are as follovs:
'j1 1.ength of stub tube. Radini force par.'irch Maxhaum band l
in. of average circu:afarence, stress, ib/in.
, Ib/in. _.
Infinitc 91,100 -388,000 5.84 91,100 388,000e* 3.89 92,200 398.000 1,95 99,200 498,000* 6.974 316,000 2,870,000*
%, f ,
- Calculated at a distance 0.974 in, from weld, not necessarily mariansa strees.
OmCK> . . . . . . . . . . . .. ....~...$.*..... SURNAME > . . .. .. . . . . . . . . . .. . . - . . . . . . . . . DATE > . . . . _ . . . . . . . . . . . . . . . . - . ,
#brm AEC.818 (her P-63' u wwmemm e au m
=
( ( The table shows that the force and stress depart signifiaantly tree the values for infinite temsth only when the tidee is less them fear saches laug, but that for lengths under two inches a redisal increase is faeree and stress oceurs. j % 4, . hA The ensimusa bending stress, tensite on the outside, is spyresiastaly one tank
- f. ram the and of the tube, or at the shop weld if the length is less than eas.
inch. ( pther studies The stresses is the field weld here been asylered semi-guentitatively sad aise the effect of shaar forces on stress and strain la very short tubes. Bothing was found hot easts doubt on the eredibility of the sendition as described.
. If, however, part of the distertions were eeused by bet plastic defeer,ation, the stresses would be lower.
I An attempt at analysis of a stub tube of wasymmetrical length was aheadened as being too cumbersome. 1 1
-i l
t I e i 6 I sumur > .. . . . . .. . ... _ . _ _ _ . . . . . . . . . . ... . . . . . . . . . . DATE > . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . rm AEC 818 (Rev.F83) u s.sownu wTmmTms emcx e-ais sa
. .* w.'.
, ..a+;
e
%, *-lvy qq Uff'*I . '. t.
i + 9
* -4
,q, g.., , ..
C . . . t =<, )~<
- R.] *p
..m ..sOR.c ,W9 ~ , A,Af
_ <v
- fp l 4*tf Pi
.a a. yl W; ;lR. .. , l' > . , . E' i ' .l ;9.p$%t_
(,. ,. [ , a ,ag *. %g L , ;. _.n .
' ', . ' - .A + = . s .; 7.'1Milh*g;p. ..
pp. "/ e, , swq x . sqr*4V4.r - l7 l'
)
s, l 3f-l! o p . y;
+
y
- 4 q ' ,(y. )
- - O.02 e /3xft c3 j) x. +)j&y ,) -
4 >. . 3 ..
. y f' *,.:]&. - -
i jy, 't' s h v 2/
-a vi i .J-4 y 4 4 . .+..
/ ;. 4 \
\
t / e
\ m*
x .' .
+ i
- 4
'/.' .'s
.I,) ;
. o 4
4 \$xy
,6
~
-j 7 g, 4 -
'7 ~ j +- ~~- - " + - ' - - - * - ~ * '~ -~ . ~-~ ~ % Q C G _
( . ~ . ~ < - 5, p.3 14 -__
. l l 2. 3 1.',cl':> ft. ,
,. . . j .. .. .,
c - c.... .. . . z, y T $ 5 Ur /k ,' .
.-..z..'.--....-.;....,; . }
i
\',
N _e. 8
# a s.. 'T 'j. E i- / \. e c:, e /> r -
s
- s. :
',o-
- *S _.. h: : .
.w - . . - .\ u s ).a.. ~ . .
.s;: ': .. ',
i f ,,. 3 i t V 2',... , '\
- I
~
! \
L k k
,-a ! !
y~,,,,.-' L. ; N I ,
. _ _ , / ,
_s\ , m _. . o /
/
i 3 2. / O l 1. 3 /pgr3
- f. b . ' V ~ .ELb' l bElo W VE L D Fic. 2 L?b li' . ? !' 5 LALC 4^~r:b S T.7 i ! )
I 4 D J ' ] A '. ,.' *A 11J i .* v' t, ..
- . p
,1
- hYl..h i i, .
A= *t .. . . - M W W4-* e
o-o .
, k.
4 ,QL4p p kw, L ..
. . .es v , . s.. ,' 1. o. , .
4 m..,
,e .5 -v tg 4 Itg&.,..v e ,, e ... ,..q.,,.
. t:
. . , . ';j ,,%, 4 ..,/. ,, A, ,./. _ .1.-
.-.-.9..,,,.7.,,y,~..y s . ., ; ,
e m w.,r.- ,,<. en
+
' ,%.V. : vn .,;
s
,e.i
< , w' e
e
. . ,' :: ;.. .P ; a. .
., - , . f r- -
,* i i - e
- * . ?AMv, , .. %(g.:,t s -
(,;1 * . I ' N.[/ )2t >' er6 , l t l 'Q.4 f k e o yrh~w'Th s j
((/- . ;
- , e ,
% J 6 8
- f. . . . .
. .. e s-. o
- ,!.. * '. . e i. ..
L. ._' 2 " 'j " . _4,.._,__..__...._4 .-_-.....____.j,,_.. '
, ' l U l '
ff, $ Ff f f t. 2 " *l,' 3
- i. . 3 4.
*s- e "9. P e t.
h. a O a' I 9 mw' 9 w k i l
- . . s. -
gr% j ' W, %*\% .% S %
. ,- --..#- .' --~ --.. .._ . y
. . + . -
a v, - . . 9 . . .-.% i e V'5 i
.)
+ 9 k'
* ! h
, . ., l ~-
v.L s p O
.'.) &-
?*.
. 9, I
)i. c) wm t
Q '.y %--.----. n
- T
. yl r r ..p yx f, y v 7, _c Y~ J + ~
k t, f *
- b
- 0 -
+f - ~
w ._ -
. . s ?'*- . e' e 9 ,
. _ . _ _._ . . _ _ _44,
't, _-
N. N .. . , .
,t ' u r: s) '
- 4. , ?
. .EEPNW 4 w'*
C f,
+&tyt ?;'ytty@t
( 2'.g' I ' g fg8 7... . ...
~* -"k}h 4
' ' i 4- .
~. : awe,4m ;
i } .'.h.hi.,. . ;8 h,
. ; . c, 1 . - . , . ,3.,.'q;} . a.
1
) ->ili';f ,
3na:p.-py
-- . .4..
g'
' _ [.. 1
( ' # h #
- ~ ~ -' . y ,
y . s 7 ; ,e f.J. d. ' " I
- q u.y'J
\ % 1 YG f, h *s e .
4 < N *L .
,= . u -m '
* ' * % A' : (, ,
s-q 6 s,. C . *^ U-' $ $.d* ) *' ') % [.~(; -
\ /. * '# ' I 3 V3 7'S' /.. [ / A' ,
# 5 g ,
i # \'
\'
i, 8 y .
\x.s i
%, y
% .. ** 4
' ~ ,
' b~ '
.-. .- ] . , . _ . . , _
sy _. ( --
-- . T '. a. ..
S 0 S. e 0 5, g3e . - . - * . A- 1 w . v
} l } } , , , ,
.,<-..,. s.,.,.,,,,,*.
$* ]
** 4 g
[ h-+ee =+we***==% use e- ww g ess ,, vo--.*g..
- m. - -,
.m
,me, h 5 Mi'N. wa se h
~.
- I h
'e w
4
- m. _
s-s- V'
*
- s
. s 4
9l 'k 5 l t . y- - s j \ . k ' tr.- - t e
- s. _ . ..___ _ _ _ _ _ _ _ _ . -
'" f-f- i M-T[
i [ . $%.%..
%s- m.
x i p 2E M e m f G - E 7 4 8888. 2 I!B W M F R N M M *
.e2 s this seg as taas enumetre them that ser the fiaM seM h 4R 83 andersesse that a detailed stress analysis has been made ny es and tiefstems the eatenstated stresses are emailer aban these deem he
,sgg, . -; . ;y su % , > w . 9,a bhb a patwo of Streep ,
.,Q
, g: ,,
b w&;p-indesed stress is nameed by diftarestial esetzestian af the steel ses see and the yearlit.A. steal weseal (and temasst esM) durias coolias i from the stresa emanating operettaa. Suppose the vessel emi semb tube age * " essantially stress free at ther stresa estiarias enspeestosa. 44. ee64 es refinement of a11 swiss for relatively esalt sosident stesesos da the toes - materials and the latter stresees prehens in the nosemal eeM. If the eenA , tube and vessel sure free to eastrost $sespondently, the essaide medima of ,, the stab tee would be 4.010 is. amm11er them she surrespositas apostas en a, the vesse1.., the and tube is therafare pallad out es sheen la Figure 1 & p+, - l1 N behavier will be between tue aarremos: . I
- 1. the tee may be hem riaW1y es la Fleure 1 b. -
3
- 2. The weld may yield and rettsee the bendias mammat as la Fleuse 1 o the actual behaviar, far a ste tube af given dimeestems, will depend sa che strength of the weld and the length of see tee projesting em seek side. If
- f. the walk does met yield, or if severet insbos of tube projoet en asek side,
- esadition 1 b preystis. If the umid yields, ubish is more likely Lf the weld i is aballow or if saly a short laestk of the protrudes as saa side, thes somdition 1 e prevails.
Case 1. Stress ta st4 tube with tia u Osastreist. per the asse shown la Figure 1 b, seastants of the gemaral faunnsla of Atteshment & ase & = 8
- 0 and c - D - 8.01.
p
- 22,800 lb per inch of strc.meferesse at .14,100 ta. Ib. per task of stroumferesse.
j tiarimias stress - 150,000 pet (at the weld, and tensile en the estar surfase). In addittoa there will be a hoop tessile stress of abast
- 13,000 pai.
I <.,.
- For a tea projecting at least a few inchas sa sesh side of the meid, the eco11tian is the sans manapt that p is asablad. s 27, ,
l
! Strese in Stub Tube with Free hd. per the ease Aswa la Fissee 1 e, saastants
{ . of the pasral fanenta are A = 3 = 9 = 4 med 8
- d.91. i f i F lb per fash of strewnference. The naziness beesla Etress k H -
psT 84.9.9741a. from the weW and is compressive at the supeset ; h M MI' _ . . _
. 33.. _.
3 Oh
... c.
su m = . ..I... .. . . . . . . . _ _ _
~'" M'
'S b,.. ~b l'*I'
>'p.
N A W 18IE*' M oATt a. s ee e.,.m . _ . u.s .ona.mrems.s wruz aoww:> tie 42.
-.,._._,.,,p.
r, ,ye ,. t weews t , 4
, , - - - - - ~ - - - . . , ,-
.y __ .. _ _ _ ._ .
. _ _ . _ . . _ _ , . .__._._._.__._.,_=_._.m._... _ _ . . _ _ .
4 ( ( Intermediate Casas. Stresses for esses with only a short length of ete sabe protruding from one side of the eeld, and esses of asymmetric bending beemase of us1d yielding are easily s=1culated, but will not approach the high sessene - levela sameed by the f1*1d vs14. pf.g.
.eid ..ree.ch. n e .aai b dtog strass ta en. . u -is given by - e syws ,
where d is the ef fective depth of the weld amasured parallel to the sie (.abe axis. The direct stress is F/d, and the maatmum taasile stress is the sus of t.be two components. At the high side of the weld d appears to be abeet two f new. At the low side, allowing for poor support by the filist, the effective depth may be j about one inch. For Case 1, likely to apply to the high side, the maximan teasile stress sa
-} the weld is 32,500 poi, which is well below the yield stress of Incomel -! (55,000 psi) an6 just below the yield stress of the statalees steel (35,000 pai).
For Case 1 with a through tube, the tamails stress is 13,000 pet. l At the low side, Case 1 would lead to a mawi== calculated taasile stress of 107.000 psi, and the ta'alens atml (and possibly the inconal) would ytald. l,. The condition vill therefore shif t- towards Case 2, with a direct tensite stress of 11,500 psi and a residual bending =a-a=t. Direction of Intergranular Penetrati_on. Case 1 of Attachraent C approximates the high side of the shop weld with a short protruding and of the stub tube, I The bending moment is sufficient to cause yielding in the outer surface of I the tube near the weld, and may bo _ sufficient to sause yielding in the stain-i , less steel alona the weld interface. This would surplain division of the crack into two branches with one parallel to the weld interface. l The above description applies equally to the bending =amant caused by the field weld, azcept theit "may be" becomes "will be".
+-
~
l : f t f i O MICE > . . . . . . . . - - . . . . . . . _ .
... M ,.11.. ?. . . . . . . . . . . . . .. . . . . .
SU#NAME k .. .. . . . . . . _ . _ . . . _ ~ .. . . . . . . . . . . DATI > , . . . . . . . . . . . . _ . . _ . . . . . . _ _ . . _ , . _ , . . . . . . . . _ . . . . . . . . . . . . . , _ , . . . . , , , , , _ . . _ . , . _ _ _ _ . . _ _ ..
, _ m .., _ _ _ _ . ,
7aWWJ/ffN g p7((a C lh df{Q ( h. sJ MAf( f@{$id{s}$y.tga%I .6 w ;, p ,;. y.7,4. v g
.. -. , . g. 1
. ,.r-e' .j, ~;' h,(;$ y i kJ3
- s;a . z:t4s. .r;s 5'
,* I.k[.I. F h,,
~
,e' ,
['
& -( -
;6fg -
.9-
/ . :. . a 1 k s4 4 e p- q,+
=
,. c .
\ '#~ .
,e, d - .
) 's '
6-
'q;r. Js .:'
- i. L ,;; ' \
\
. .- s I .p:.<n/
s .
; . e-
.F, N '
.T s N jT e.
L b ld -
.~
j l l~: Face ie . e 71 % b - -. P A
. : FsA /6
_ -, o - _- - - -
, >x l
-I u -
-i -
( a
*
- I f $ {t b 7
p E E s f } $r' f k5 T) $ kt? 4? *'E k - '. '.;
,f f g . .
4 I$' - l
- @r
~,.
- h !
1 (1
_ _ _ . . . _ _ __ ___ _ - _ _ _ - _ _ . _ _ . . _ . . _ ___........_____.__.__.__._..m._.._.. ( (, ob# ; s > EXCERPT FROM SLM1ARY OF 98TH ACRS MEETING JUNE 5 8.1968 MEETINGS b'ITH THE DIRFCTOR OF REGtfLATION 4 AND THE REGtlLATORY STAFF r r )
- 3. Ovst er Creek Nucicar Power Plant Mr. Price advised the Committee that the applict has requested authori-zation to proceed with the repairs to the two .. tubes that had been i
reserved by the ACKS for further analysis if such analysis was deemed 4 necessary. I Members of the Regulatory Staf f reported on the progress of the repairs to the Oyster Creek vessel and the measures proposed to minimize the 3 effects of a f ailure of the sensitized lower shroud support ring. l The CommitLee concluded that it should not be involvrs in such " day-to-day" l decisions and agreed to leave this decision to the Regulatory Staff. Com-l cents of individual members (ttr. Etherington and Dr. Bush) were provided
- j to the St.aff (copy attached) .
i l { ' I, i l , l' . l , l l
- - - . - . - - - . . , - , _r ,- ..%-,,.w
< . , - . - - .. - , , , . - r-- ---,,-.g -*
.
- v vim c,
( ( ,, 4 % iC os
'I e*
Sr
-c o S
4 Oyster Creek Stub-tube Repair @q v4 I3,99, a
. e ACR3 Interest in Repain w* 4 ?
The Applicant contends that no safety problem is involved in the repair. 7ne Co==ittee should perhaps first decide whether it concurs. Co=.ittee interest could arise fro: consicurations such as the following:
- 1. Possibilit: of in-service deterioration of stub tuber, OxD housings, or field welds eight be construed at causing a credible hazard. For example, simultanyons ejection of a sufficient nu=ber of rods to cause failure of the catcher tight be concidered credible -- this highly improbable condition could be postulated on the basis of undetected general deterioration fellowed by a severe pressure transient or earthqueske.
2 It ci-ht be considered desirable to cinimize any riek that the catcher rn: be " tested" by a single control rod ejection.
'.,. Failurec in service that cause leakage =ay place on the AEC the burden of n difficult- decision concerning the safety of continued operation.
4 It ci-ht be censidered that quality assurance of the reactor vessel is important even where screty is not an issue. l 'he fellowin- discuscion is based on the supposition that the Com:ittee l' is interested in a critical review of the repair procedure. Field-weld Contribution to Stub *bbe Crackin , 7ne reacter supp1Aer hae concluded that the observed inter ranular crackin-l of the stub tutes is caused by sensitization of the caterial, a corrosive i a ent, and strece caused b- the =anufacturin~ procese. "he underscored ! r sumption in cpestion6. '. The supplier cites calculations of ! c.s-=anufactured stress by th! ee independent organizations, but is silent on the cuch greater etress imposed by the field weldc.
- 1. Has the reactor supplier calcult :.ed the stresses caused by the field weld? Does he recognize that the calculated tensile-stress from the field weld, at the loce. tion of crackin,g, may be an order of magnitude greater than that from shop fabrication? Did he censure the relief of vtrain on cutt-ing the field welds before he established his repair proceduret*
1 e
- So far as we knou, the only ceasurements to investigate the effects of the field welds were made at the request of AEO. The Applicant has been requested to measure the relief of ctrain when the remaining field welds I are cut -- it is possible that the measurement: willbereassuring.((~l-dI y . - . , ,, .
( ( b 2 The limitatione of elartie analysis and the uncertaintier caused by highly localized deforsation are obvious, but does this justify complete ciarerard of potentially extre ely hi;h field weld streseect
- 2. Invectirations have failed to explain the severity of the inter-granular attack. Have the metallurgiesi investigators been alerted to the possitility that stress and strain at the location of failure may have been evtraordinarily hight 3 The inter ranular cracks show a vertica; cranch parallel to the Inconel weld interface. High tensile stress in this region is indicated ty calculation. The stress will be increased ty removal of part of this weld. Would not this be a rencon to pay some attention to the stress at the base of the stub tube caused by the field weld?
Si rnificnnee of Fabriention and Welding Stresses. A steel veccel with no austenitic inserto is essentially stress-free as delivered. Oa culationc of operational etresses are therefore normally based on c. s tress-free vessel. However, the C,ne does not prohibit use of austenitic oarte in a ferritic vessel, and so:e level of residual stress is therefore implicit 1" condened in an as-delivered vessel made of materials havint different coefficients of expansion.
- 1. Does the cunplier take the po ition that Code silence on re:iduti ctrees i plie: that stresses produced by field welding (accu =in:
no sencitization) are unimportant, re ardlees of uncnitude, and that rood practice does not require anv effort to diminish the:7
- 2. -If a shrin3 ring were applied to a tube, would the supplier ibnore the resultin atresce: (a} in the rin;, (b) in the tube under the ring, and ( c ) in the tute awn; from the rin;t 3 If the supplier does not take the position that welding and shrink-rin;-stressec are totally unimportant, what are his criteria for acceptable strecs levels and how has he applied them to hic repair procedure, i.e. does his field welding teocedure reflect a considered attempt to mini =1:e weldin; strecses and if so how?
- 4. Has iba supplier conducted a laboratory development tect, with the actual: configuration and catorials, te develop a procedure that will minimize radial contraction of the housing and stub tube?
Use of *!andrel in Field Weldint The initial field welds have caused radial shrinkage of the CRD hounin e, and further ehrinka;e durin~ welding is anticipated by the supplier. He specifica that a candrel is to be used where the over-all
. ( ( .
. s Where the shrinkare would interfere with insertion 6f thermal sleeves. mandrel 16 used, the ctress condition in tne CRD housing and stub tubes wil1 be alleviated, but the hoop stress in the weld will be increased.
~
- 1. What is the justification for trute force constraint during weldin-? 'ihy i should a procedure that so creatly affects the stress distribution be optional?
- 2. How doec the procedure differ in principle from welding two bars ord-to-end and preventing shrinka e by rigidly clanping the barst
' iou'.d the supplier consider this procedure acceptable?
Control-rod-drive Housin s Calculations fron measurements of the houcin; deformation show (elastic theory) nn extremely hi-h tensile stress incide the housing close to the field weld, ir a re-lon that is airo heat-affected.
- 1. Enc the reactor eupplier eniculated the tensile stress inside the houcingl
- 2. Han he sectioned a CRD housing and examined it for sensitization and susceptibility to inter 7tanular corrocior.t
- 3. Did his procedurec adequate 1" protect a-ainst pcecible overheating durin: flame strai-htenin;, and has he inspected housings for any evidence of overheating **
- It ic stated that the housin e are kept below red heat by visual observation. However, the unir.hibited tendency of the shop man faced with a larye amount of strairhtenin; is to hee' n small spot to a good red heat. In torch heat.ing, it is poccible to hs ;c well above incipient redne:s (even to dull red heat) wi',hout seeing s elor until the flane is turned off. How is the operation monitoredt Conclucion
-It io probable that the repaired stub tubes will be in good condition metallurgical 17 and that inte rity is reasonably assured arsinst any level of strece. However, the lack of curiosity concerning the role played by field streso in the feilure is disturbing, because these same stresses will be imposed on the 30SL overlay. The reactor supplier apparently takes the position that weldin stress is unimportant, retardless of magnitude, and is making me effort to control such stresses. This does not see= to
_ pt
+ L,
( ( 4-be consistent with a good quality assurance pro; Tam, or to rer16ct the consideration that thorough in-service inspection will be virtually imposrible and that the proble: will apply in soms decree to the entire GE product line. De possibility of trouble in the CRD housing should also not be overlooked. H.E. 6-3-66 l' l i l 7 t 1 l l
. . . _ .}}