ML20098B121

Residual Stress Analysis of Peach Bottom Unit 3 Core Spray Pipe to Sleeve Fillet Weld

ML20098B121
Person / Time
Site:	Peach Bottom
Issue date:	09/30/1995
From:	Marisa Herrera, Stark R GENERAL ELECTRIC CO.
To:
Shared Package
ML20098B113	List: ML20098B112 ML20098B114 ML20098B121
References
GENE-523-A101, GENE-523-A101-0, GENE-523-A101-0995, GENE-523-A101-995, NUDOCS 9510020199
Download: ML20098B121 (17)

Text

q..- .

H n  ;

. N.] GENE-523-A101-0995

\

j l DRF # 137-0010-8 3  ;

w 3

5 a

y 4

Residual Stress Analysis of the Peach Bottom Unit 3 h Core Spray Pipe to Sleeve Fillet Weld

.3 ou

-: September 1995 m

t

/

?:

6 Prepared By: d k

" Randy Sta&, Engineer 2 Engineering & Licensing Consulting Services Y

.1 P

Approved By: N l parcos Herrera, Principal Engineer a- Engineering & Licensing Consulting Services GE Nuclear Energy J 175 Curtner Ave.

San Jose, CA 95125 0

~

9510020199 950928 L. . PDR ADOCK 05000278 P .PDR

._ - . _ . . . . - . . . .~ - -.

M

,Q GENuclear Enerv GENE-323-A1010995 DRFo137-0010 8 l~ '

g A

H IMPORTANT NOTICE REGARDING P5

~

CONTENTS OF THIS REPORT

!r

(

& The only undertakings of the General Electric Company (GE) respecting inforn.ation in a

this document are containedin the contract between PECO Energy Company and GE, and nothing contained in this document shall be construedas changing the contract. The use of this information by anyone other than PECO, orfor anypurpose other than that l

' for which it is intended under such contract is not authorized; and with respect to any unauthorised use, GE makes no representation or warranty, and assumes no liability as to the completeness, accuracy, or usefulness of the information containedin this h

document, or that its use may not infringe privately owned rights.

H .

1 p

M l'

b l l l W l F  :

El )

" l 6 .

h%

pan 2

W4 LtJ i

A

a. .

GENudnur Energy GTEE-3.23-A1014993 DaFn131.oot u a

TABLE OF CONTENTS E

. l' l. Introduction 1 foi 2. Analysis 2 1-

3. Results 9 r

W

4. Conclusions 13 2

p

5. References 14 m

=

Wiii L

.~2 n,

i j

u ,

J 41

!.a

u ..

t m' 1 l

jrt i me e

e

'! fIif 9

$ '4

-. J.

i B

J

(

7 .

I cex-i ee-a cene.nui0s ms DRFM13%001M

e 3 1. . INTRODUCTION h

f An indication was observed in the Peach Bottoni Unit 3 core spray pipe sleeve in the vicinity of the fillet weld. This fillet weld willinduce weld residual stress and an area of sensitized material. The residual strese due to the fillet weld is a key parameter in M' determining the potential for crack initiation in the core spray pipe. Due to the uniqueness Mi of the weld geometry, a finite element model was developed to simulate the residual stress ,

d'

?A behavior on the outside wall of the core spray pipe.

The finite element model was used to estimate the resioual stresses generated in the weld rnaterial and the base metal as a result of the welding nrocess. The numerical evaluation P h of weld residual stress is difficult due to numerous welding parameters and physical characteristics which affect the solution. Some of these characteristics include non.

linearities such as temperature dependent material properties for both the weld metal and ,

the base metal, highly complex heat transfer mechanisms which occur in the welding Pt tia process, elastic-plastic behavior, and path dependence of the solution. Thus, this analytical evaluation Should be useful as a qualitative tool to evaluate the residual stress M-ca behavior at the location ofinterest.

Y8 u- The use of analytical methdi imilar to those applied in this study have been used g previously by GE (References 1 through 5). The results of these analytical methods have d also been used in comparison with test data to demonstrate the effectiveness of the

, methods. In these cases, despite the complexity in modeling weld residual stresses, the analytical methods have been reasonably consistent with the results of the test m measurements and observations (References 1 through 6).

MiO a

1 o

.Y cesua.a t,,-ar cess.nusol-om DRFM 137-00168 l

~

"" 2. ANALYSIS l

The analytical determination of weld residual stress is comprised of three steps. These steps are:

1. Finite Element Model Development

2. Thermal Evaluation of Welding Simulation u

3. Elastic-Plastic Stress Evaluation i

-a The finite element model is developed to approximate the actual geometry of the welded joint between the core spray pipe and the connecting sleeve. The thermal evaluation provides the transieht temperature history at each node in the model for the simulated weld pass. The elastic-plastic stress evaluation then uses the results of the thermal analysis to determine the residual stresses. Each of these steps are discussed further in the y following sections.  ;

iL P 2.1 Finite Element Model L

J i2 A two-dimensional finite element model of the cross-section of the welded joint between L...

the core spray pipe and the connecting sleeve was developed for this analysis using the ANSYS finite element analysis program (Reference 7). The modelis composed of1522 l

axisymmetric elements configured to closely approximate the actual weld joint geometry. l The model is shown in Figures 2-1 and 2-2. The simulation of the weld pass requires both x[

a thermal solution and an elastic-plastic stress calculation. This procedure is described in re

[ detail in Sections 2.2 and 2.3 of this report. .

I 1

L 1 All material was stainless steel. Temperature dependent material properties are used for

, both materials ranging from 70 to 2700 F. A constant convection coefficient is used for l

[ all outer surfaces of the modei. The top and bottom surfaces of the model (see Figure 2-

.. }. ,

2 I.} i l

,a .- - - . , a u- - - . , . . . - -ana. , , -- . ... =x=.- a.-.--. ,.a.. < , . .

jm . . .

A GENuesner E=gy GENE-333-A191-0995 4

DRFG 237-00168 i.

1) are extended a distance of approximately 3V(rt), where r is the pipe radius and t is the

, pipe thickness, to insure end effects do not affect the solution in the area ofinterest. The b thickness of the core spray pipe is 0.280 in, and the thickness of the connecting sleeve is 0.219 in. Elements at the bottom surface in the model are restrained from motion in the l--

area ofinterest, and elements at the top surface are coupled in the y-direction. No other restraints are applied to the model, and thus it is free to deflect during the weld simulation j

process. The weights of the pipe and sleeve are not considered in the analysis. Therefore, Ig the predicted stress profiles are comprised solely ofweld residual stresses.

a-i m

d 2.2 Thermal Analysis 1 a 5

A thermal analysis of the weld pass was performed on the finite element model described 3

in Section 2.1. Isoparametric thermal elements (STIF55 in the ANSYS Library) were

?

used to determine the transient temperature response for the modeled weld pass.

3 ,

i  :

5 1

d The physics of the heat transfer near the weld pool is highly complex and requires a computationally intensive and highly specialized analysis to model accurately. The a complete simulation of the heating process should account for conduction and convection

/

p within the liquid weld pool; convection, radiation, and evaporative losses at the weld pool

' M' surface; as well as heat conduction into the solid surrounding the weld pool. In order to a simplify the modeling process of these complex heat transfer mechanisms, GE has employed a modeling technique which can be used to calculate the trends and approximate 9 magnitudes of residual stresses generated by a multipass weld. This technique, called the nugget area heating method (NAH), simulates the weld procers by heating the nodes in P the model which lie inside the region of each weld pass to above the material melting d -

temperature, holding this temperature for a specified period of time, then releasing the LQ imposed temperature constraint and allowing the material to cool naturally. Each successive weld pass is simulated in this manner. The area of the heated weld nugget, the

] time required to ramp up the temperature, and the time at which the nugget is held at its 3

i

9 GENE-323.A101-0993

,d GENucler* Enerst DRFM137-00104

$ maximum temperature are determined by the welding parameters. Figure 2-3 shows the NAH method schematically.

r-r To simulate the weld pass, the material is heated from 70 F to 2700*F. The amount of time required to ramp up the temperature to 2700 F and the amount of time the weld r nugget area is held at its maximum temperature are based on weld travel speed Y parameters.

b_

f)

'5 .

E3 r

b /

1 I~

a

~

I Y,

I b M

-w

=

med e

.J 4

I a <

La IE'E] 'lt m IEi] il u .; iEc TEE ILa i~ED i bi N II9 T _tj 7 .N 7' 7 ,_ m 1 ANSYS 4.4A1 -

SEP 25 1995 14:23:54 PREP 7 ELEMENTS TYPE NUM f

ZV -1 k DIST-4.813 XF -3.298 N(

YF -4.375 m

. Gi-E a

P

.T 21

=

=

o M

\ ,-

E a

E E.

a._

j t

,a

~ :;;

t t>a V

7 X Peach Bottom core Spray Residual Analysis _ _

~ . .

ti f' GENuciner Emrgy GENE-32M101-0M5 DRF E 137-0010-8 k

1 E

e 1 i

I r-1 li$

k N to

\

H

\

b

, ._ N H \

Ja 1 ,

..x l

l

s. '

b

., s a@a i

i h -

d

/4

,Q

!w

=

l l

,a 1 5

I l

a tt

,.2 ,

. i

}j Figure 2-2: Finite Element Model- Close-up of Weld Joint e

a 6

I

3.' ' l a GEN: clear Energy GENE.32341010995 DRF0137-001M E

E i

.t-1 i

1 E

i Temperature g Temperature i L  ;

/ Boundary

, Conditions g;;, ',

Released g . ,...'

h 3 -

Time 3 Nodes 1

0 '

A / g h

\ / '

ry '  %~ ,r" Weld Pass N Boundary

%e.7 l ,,

( -

l

us u_

m

L

ta i-m, q$

jj. Figure 2-3: Schematic Illustrating Nugget Area Heating Technique

i.

6

l7 GENucione Emrgy GENE-323-A101-0995 l DRFM 137-00168 ,

I n

l$ 2.2 Elastic-Plastic Stress Analysis I The elastic-plastic mechanical analysis uses the results obtained from the thennal solution a to calculate residual stresses.

m Since the heat generation due to plastic deformation is negligible compared to the heat

O 4 flux imposed by the welding process itself, the thermal and mechanical analyses are y-y performed uncoupled in this evaluation. This implies that the thermal and mechanical 6 analyses are performed in series, as discussed previously, rather than simultaneously.

The mechanical analysis was performed using the same finite element model developed for the thermal analysis. For the stress analysis, the isoparametric stress element was used

} (STIF42 in the ANSYS Library). The model assumes that there is ne azimuthal variation in the stress due to the weld. Although in actual welds there is variation of residual stress

} around the circumference of the shroud, analytical methods have been found to give h

reasonable estimate: of a typical location (References 1,4, and 5).

! Temperature dependent material properties for the clastic modulus, thermal expansion coefficient, and Poison's Ratio were used in the stress evaluation. In addition, the non l tt2 gj linear properties ofyeld strength and tangent modulus were also input as a function of temperature. The matedal density, for both the base metal and the weld metal, was g

ij assumed to be constant in this analysis.

M d The required iterative solution process ends when the user specified convergence criterion is satisfied. Plastic analysis is performed based on the Von Mises yield criterion and the e

3 ,

Prandtl-Reuss equation. Subsequent yielding is evaluated using a bilinear kinematic  ;

w hardening model. The final stress state of the welded joint is obtained when the solution 5 for the v/ eld pass is calculated as the temperature in the model cools to its original steady y state temperature.

J c

8 d

J

GEN: clear Eurgy GENE-323-A101-0995 i f DRFM 137-001M

.9 3. RESULTS

. c

" Stress contour plots for the axial stresses, oy, for two cases are shown in Figures 3-1 and T 3-2 respectively. Figure 3-1 shows a sleeve-to-pipe gap of 0.0625 in, and Figure 3-2 t1

^~ ~ shows a sleeve-to pipe gap of 0.031 in. These plots show the stress state in the welded

-i~ joint after the weld pass has been completed. The scale on both figures gives stress values M . .

m psi.

10 Figure 3-1 shows compressive residual stresses on the outside surface of the core spray j pipe in the vicinity of the Heat Affected Zone for the 0.0625 in. gap case. Significant tensile stresses are predicted on the inside surface of the connecting sleeve, which is consistent with the observed crack which is postulated to have initiated on the inside h

surface of the sleeve at Peach Bottom Unit 3.

m ,

Figure 3-2 also shows compressive stresses on the outside surface of the core spray pipe

(~

=

c in the vicinity of the Heat Affected Zone for the 0.031 gap case. Again, as in the previous case, tensile stresses on the inside surface of the connecting sleeve are predicted.

l l

l" ]

l

,r Figures 3-3 shows the hoop residual stresses for the 0.031 in, case. These results indicate Lu compressive hoop stress on the outside surface of the pipe in the vicinity of the crevice.

i-8 n4 4

b W l h

9 A

<T hfl_.f H~Lc f ITl'2 EET hi IEif r t E ii~-~ M 'i ~~ T eid ' Eni - kdl . tile EEd T3 E E l 1 I ANSYS 4.4A1

• SEP 18 1995

, 18:18:43 POSTI STRESS STEP-342 R ITER =15 j SY (AVG) {g -

S GLOBAL p

(

DMX -0.003927 SMN --42675

.yc SMX -56809 5 , ZV -l Y

- d *DIST-0.565335 sXF -3.32 '

M

• YF -4.506 2;- i '

=. -42675 f-' '

5

~

l -- . -31621

[ i 5 77 -20567 d.

ti i ......c -9514 l I  %! ! , .. i ~'

1540 i

& &l 69BilR@.s iM$tEd j 12594 r-i m 23648 j

e M ninMMMRiM@ Ele i i i- l 34702 3

2 W. ~T.!h!4%-

n l m ,

56809  ;

D -

x i.

w I

,8 o .

s

' kE '

E!=  ;

w t

?

l Peach Bottom (,, . . hal  ;

i

__. . _ _ _ _ _____ _ _ _._ _ _ _ ___ _ _ _ _ _ _ _ . _ _..._._ __ _ __. _. _ _____ m.. _ ._ __ _ __. _ . _ . . . . . _ . . .

O L-- I E-~1 La I La i u a to ta < Q~ em LA i CEi b .h ;- cal ETl EU EE c a.1 i

1 l fig ' ANSYS 4.4A1 '

f[+lj j'ldq;j. Ik g

gij l

SEP 19 1995 11:55:13 d li di&ggf ns li_ POST 1 STRESS o i

I5}f STEP-357 h

"~I3f@tjg i T li

-d p: ITER-15 (

[i A ;'p SY (AVG) gi

, L-S GLOBAL g

[.-/'I:-4 5 DMX =0.405G86 h g I[IE SMN --7J650 y {4 8 d SMX -47404 ta L

• su

.h . . .a ZV -1 g

a 3gp

'[ f-f=g

?

• DIST-0.53276

• XF -3.313  ;

[ . N,, E .Y E ,.

• YF -4.482

'2m -76650

$ f4_

E. 4_43

& M4_+_4, hjdt)h - -. -G2866 g- ravy e e i Mkt 3 4 i wi r -49082

,~  ; T'T Ty 4,iE ]f;i# ^

~IfM{

J .? -35298 Ou) fT T T T ;(~$~ [f:T ~

~

-21514

.:+-+-+j,h- i + -.g - -i , -7731

$  ! ' E ! .fV h'E-4[ , ,

6053

!  !!kh m h.h-t-+ E 3$$N 9 ._

.h .

0 1 l l l[fh$ i _&% if[ en Ll 4 %;!$lR l

li$ 'tihL '

h$%

N::a

$l  % led '

lie T. T i,g15I wa

.~ iwi AE

!I l l linIS:?6' ,e '? Ia6 ,

T T ; ; a : ,t.d' P a:

i: iT2 i

i .;

')

l+ '

c- .

'+'+h@.!

l l l l

.g

= l l Peach Bottom C4rb Sp)a. [, J

, sl(,  !

t f

--_ .. .. - -. -- - . - . . . . .-~ . . - . . . - -.- .

GENudar Ewzy GENE.323.AIOM995 I

,t DRFi137 001M l

l

, m-w I

, t=1 La W

. M 4 M M W

! 4 lin M W CQ Ln CQ N 1.0 c) M N

! . c1 W ^ Ln M c) CD M cQ M M N r=1 r=4 CD W N C1 9 c) g C C3 N e-4 cQ Ln Ln N N N N N N C3 CD r=1 e=4

- e-t e 4 H N > J e N cc T N W N W Cn M W CD CD CQ M CQ j 4 H c1 Ln . . . CN N la T N H M N Cn Ln M N .. M cQ N HLn ( < atn vr.:cm e-ICs M e I e a I I I W N M Ln

> N c3 c-I e u oa e a aean

. M LnHEE _J H ""

- 211.**M W W CXZX M  : - 7 9 L <wNCHHN EEE >Hww 8 g j l- M A M H M M C3 M M N c3 X > -

I _.

3 i

i

<m i

d t

4 cm i

4 -

s'

%$N L.4~ f

^

-gjf fi - S T -? '- '

i?r- 4f!!f,._ _.+.Yb.}1,7 .W Y f_..:_;_ _-f -- :

m /d g !! Ib+i__ $ _ .p _ +_==w

_ _ N+' q

,.a. -_O p521. ,

p /

g_gbysigt Jgg s. _ _ _ + _ .p, 3, _ . _ _.2___p, m

w p

g W

_ Lb__ h_ -C \

$v a% ka%+ _-_. _ l_1_'l _1i. .t _ _ _ _ L fit a I? t e n& % ysi m.t.,,%Lf  ?._ , ,

qifg,fgath.W

_m _.7 y 44 Q;4@ Rs %.44 w

% w_ waav '.u_m .

4 g U@% $hWS em. Nr*h a

..i M 5W D,t i$g f[5[63rIjh5 IjB8rY qF,sq

_+T 4 Fyeips gAA%i haw $2Eib;G%gp%g

$ %:4 .g@... r: y3

. x (a.sigigw~y mp 1 x$c$IypdhjhgM$:%g$phjfI~~~~

etEr-es r.- t -- -- - 'j"-

rf ~

g e.;grj~+3p.De c 2 ips+ gw %j ga a;hv - i h -----

n: -- g-a, 4._ ~ + v. _~ y we s pA y e.

~~ ,: 2- *"._ p. w-.n.d. g' g" u. s' _ _ p _ _ _ _ .

i 3 .iR i i

c+ s i ft@gr{%ggg+-i@g g(553)'~ys g, - tg... ..~gf.___._

g yp - - -

- t - - - - ,---g 9.

I w " f5$h $$ M  :- .Jg$Vp_ r _ l.. . _ _ _ . _ _i p . _Ii.

9 .h. *h_hh.i;2a.:.$.hk___I'~~~I"~~I'~~~

m t____.t.____A____t.___t_.e:

i_

o U

fii s Ld 0

e a

w o

cc J

.c U

c 4 o A

Figure 3-3: Residual Hoop (o,) Stresses for Case 2

g....  ;

j_) cesars,au ctxr-s2usos.ons

\

DRF3137-ooiM l El 5

4. CONCLUSIONS a

A Results of this evaluation indicate that fillet weld induces compressive stress on the g outside surface of the pipe in the vicinity of the crevice. Tensile stress is calculated on the 5

inside surface of the sleeve. This tensile stress is consistent with the observed crack y initiation in the sleeve which is postulated as initiating on the inner surface of the sleeve and progressing through-wall to the outer surface. The compressive stress found in the m pipe will inhibit the initiation of cracking in the pipe.

N Es d

2 M

h .

m vil N

l4 i l

N 2

C3

?,e e

a.

e W

-y e

13 a

p

. . , . . l

,gl GENuciar Enay GENE.323.A1010995 DRFR 137-0010 8 M 5. REFERENCES

" 1. Herrera, M.L., Lenge, C.H., Ranganath, R, Analytical Evaluation of Residual Stresses

• in Pioing Subiected to Induction Heatine Stress Imorovement Process and Aoplication

.L to Operatine Plants , ASME Technical Paper 81-PVP-19.

=

m 2. Herrera, M.L., Parsons, S.R., Chapman, T.L., Residual Stress Analysis of a

. Swecoolet Joint Subiected to the Induction Heatine Stress Imorovement (IHSD

= Process. ASME Volume PVP-98-3 M

3. Herrera, M.L., Mehta, H.S., Ranganath, S., Residual Stress Analysis of Pioing with Pre-Existina Cracks Subiected to the Induction Heatina Stress Imorovement iii '

Treatment. ASME Technical Paper 82-PVP-60 F 4. Herrera, M.L.. Analvtical Evaluation of Residual Stresses in Pioing due to Last Pass L. .

Heat Sink Weidine, ASME Technical Paper 84-PVP-36 i FJ lL

5. Rybicki, E.P., Stonesifer, R.B.. Comoutation of Residual Stresses due to Multipass

's Welds in Pioing Systems, Journal of Pressure Vessel Technology, May 1979, Vol.

!- 101. /

iG lL 6. Rybicki, E.F., McGuire, P.A., A_Comoutational Model for Imoroving Weld Residual Stresses in Small Diameter Pioes by Induction Heating, Presented at the 1980 a

Pressure Vessel and Piping Technology Conference, August Il-15,1980, San Francisco, California.

i I' b

7. DeSalvo, G.J., Ph.D. and Swanson, J. A., Ph.D., ANSYS Enaineerina Analysis g

t System User's Manual. Revision 4.1, Swanson Analysis Systems, Inc., Houston, PA, March 1,1983.

3

,=

b

, L.

l 14 d

- . . -.