ML19343D349

Fracture Mechanics Procedures for Primary Component Support Toughness Evaluations.

ML19343D349
Person / Time
Site:	Arkansas Nuclear
Issue date:	02/27/1981
From:	ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY
To:
Shared Package
ML19343D342	List: ML19343D341 ML19343D344 ML19343D349 ML19343D353
References
PROC-810227, NUDOCS 8105040329
Download: ML19343D349 (32)

Text

_ L g '- ..

Fracture Mechanics Procedures .

. For Primary Component Support Toughness Evaluations

' Recomendations for Inclusien in NUREG 0577 i

i .

p 4

a l Combustion Engineering, Inc.

February 27,' 1981 e e a

[. . 8.105 0 40327p .

uw+ww.,-.e~<y,,,-+,wr,-<w-c.~.me--,ewm,. + m, m. y,,v, ww ,-e m-, - y ,w,

• --,e-orw- .r-e-see-*+ee.og,----er,w-~ww-.,--r,,e.wyer(,,,m,ei,a ,- y rw w v-- v w g ', -w y t --. *

• c .

Contsnts .

1. Introduction ,

2. Support Configurations 'for Which LEFM Could

.Be Beneficial -

3. LEFM Evaluation Procedure for Lugs and Clevises 3.1 Stress Analysis of Lugs and Clevises 3.2 Kg Determ.ination -

3.3 ~K IC Data for Lugs and Clevises 3.4 Reference Flaw J

3.5 Acceptance Criteria

4. LEFM Evaluation Procedure for Support Skirt Geometries 4.1 Stress Analysis of the Typical Cylindrical Skirt 4.2 Kg Determination 4.3 KIC Data for Skirt Materials 4.4 Reference Flaw ~

4.5 Acceptance Criteria 5.

. LEFM Evaluation Procedures for Beam-Column Sections

6. Preservice I'nspection -

7. Summary of Recommendations

8. . References 4 .

, h G

9 y , - , . . ., --,-- . 6.,re-a- , ,4.-,,-,, . p-. , ,..a .s-.- - r, s..-- - --e. p.

. 1. _ INTR 000CTI0ft Th: ~

2 NRC report, NUREG 0577 (Potential for Low Fracture Toughness-and Lamellar Tearing on PWR Steam Generator and Reactor Coolant Pump Supports, Reference 1), was issued for coment in late -

1979.

Modification to the evaluat' ion ' procedures and implementation of the NUREG were made in two NRC l_etters,in play,; 1980 (Refen?nce 2,3).

One of the significant changes was th.e elimination of the option to use linear' elastic fracture mechanics (LEFM)Jas an acceptable support toughness evaluatio'n procedure.-

In order to reestablish LEFM as a credible method for support toughness evaluation EPRI was co:m:iss.foned by the s ut_ilities to -

prepara a detailed procedure which the, NRC would inco,rporate into the NUREG. In the latter part of 1930 CE decided to assist EPRI in preparing a justificatic r the use of fracture .

mechanics 'for support evaluation. After the Decerber 17 meeting between NRC and Industry, where EPRI made;its final. presentation, (Reference 4) the NRC staff stated a list of requirements which a fracture mechanics procedure must meet in order to be accepted t

for inclusion in NUREG 0577. These requirments include (Reference 5):

1. Specific applications in which LEFM can be applied and in which it would appear to be .

of the greatest benefit.

i

2. Geometries in which LEFM usage is proposed, including a proposed solution for each ~

j geometry.

3.

Detailed definition of residual stress for each proposed geometry.

4. Detailed list of data available,especially KIC listing by materials.

{

5. A definition of the " reference" flaw to i be used in the LEFM analysis.

en -

m

,-a, - , - ,-<n ----,~--,+n v - - < < - -- - - - . , e-- ~,w - - + -m ,r--- w=-- -

6. Proposal regarding margins of safety, especially the parameters to which they should apply and -

proposed values to be a.; signed.- - N Combustion Engineering recommends the inclusion of the.'. fracture mechanics option into NUREG 0577 and presents -a procedure responsive .to the

• NRC requirements in this report.- -- -  ; . . :.p a .

2. - Support Configurations for Which LEFM iCould be'Beneficialc . c -

Fracture mechanics analyses can be used to evaluate the:1ntegrity of structures.which exhibit brittle behavior."-Structureseare often quite flaw tolerant even in a brittle state if the stresses are well below the yield stress of the material.

Structures and supports, therefore, with relatively low stressas may~be demonstrated to have adequate fracture resistance by LEFM analysis.

Some of the support geometries which are likely to benefit from an LEFM evaluation include lugs and clevises at pinned connections, .

support skirts and beam-column sections. A typical lug geometry is shown in Figure 1. The ranges of key dimensions are listed in the figure. Figure 2 shows a typical clevis geometryr -The range of key dimensions are also listed in this figure. An example of short lugs and the range of dimensions is shown in Figure .3.

A typical cylindrical support skirt geometry is shown in Figure 4, and.a typical support coluan is shown in Figure

~

5.

Many other supports have regions where the geometry is similar to the examples cited in Figures 1-5. It is expected that the procedures for LEFM analysis will be equally applicable to these respective ,

regions.

l 3. LEFM Evaluation Procedure for Lugs and Clevises l The LEFM evaluation procedure is based on a detailed stress analysis t

! l 1 -- _

6. Proposal regarding margins of safety, especially the parameters to which th3y should apply and

..' proposed values to be assigned.- -  : - . .

Combustion Engineering recommends the inclusion of.the': fracture mechanics option into NUREG 0577 and presents-a procedure responsive.to the

• NRC requirements in this report. d 2 1 -

m. . .c4. : . ...-

2. Support Configurations for Which LEFM J Could_ be ~Ber.eficial. . I- .

Fracture mechanics analyses can be used to evaluate the! integrity of structures.which exhibit brittle behavior. -Structures. are often quite flaw tolerant even in a brittle state if the stresses are well below the yield stress of the material. -Structures and' . '

supports, therefore, with relatively' low stresses may:be- demonstrated to have adequate fracture resistance by LEFM analysis'. '

Some of the support geometries which are -likely to benefit from an LEFM evaluation include lugs and clevises at pinned connections, .

support skirts and beam-column sections. A typical lug geometry is shown in Figure 1. The ranges of key dimensions are listed in the figure. Figure 2 shows a typical clevis geometryr -The range

~

of key dimensions are also listed in this figure. An example of short lugs and the range of dimensions is shown in Figure 3.. .:.

A typical cylindrical support skirt geometry is shown in Figure l

4, and a typical support coluan is shown in Figure

5. .

Many other supports have regions where the geometry is similar to .

the examples cited in Figures 1-5. It is expected that the procedures for LEFM analysis will be equally applicable to these respective regions.

3. LEFM Evaluation Procedure for' Lugs and Clevises l

The LEFM evaluation procedure is based on a detailed stress analysis l -

= .  : -- -. -

' ~ ~ " ' '

3.2 .Kg Detemination .

The stress intensity factor, K , can g be ' conservative 1y' .

';:a computed from the stress distributions' in the uncracked

. . . . . .. r-structure by the procedures of ASME Section Xf.' ' An" "

example of this process is given~ in Figure 10.~ For' the

' ^

case where ro = 12 inches,'rj ='4.3 ~ inches and th'e crack' "

depth is (rg-rg) /10, the stress intens'ity' fa'ctor is' ,

~ ~ ' "

computed to be: '

0.429 Ksi h/(Kipload/inchthick) 4

, . .. .c-A finite element analysis of the same geometry was performed using the MARC, general purpose finite element progfam (Reference 8) with a J-integral procedure for the coniputation of Kg. The finite element model is shown in Figure 11, ~with crack tip detail shown in Figure'12. The crack tip elements are eight model quadrilaterals with mid-side nodes placed at the quarter points to represent the appropriate'1/ F singularity for LEFM. The Kg computed from the finite element analysis is:

~

.. :c .

0.373 ~ Ksi h/(Kip'l'o'ad/inchth.ick)

As another example, the case where ro = 5 inches and rj = 2.75 inches is also considered. Using'the'same procedures as in the previous case, and a crack depth'of (r g -rg )/10, the' stress ,

intensity factor for the Section XI analysis is:'

0.850 'Ksi /(Kip load / inch thick) e t_

and the finite element analysis valu2 is:

0.794* Ksi /(Kiplo inchthick)

Other geometries within the ranges describ'ed in Figure 1-3 have been evaluated with similar results.' 'Compar'ison 'of Section XI and finite element results show that' the' Sectio XI f;ormulation produces conservative yet reasonab.le~ values' fo'r' k . These -

g procedures are simple to use and are widely accepted industry practice.

These examples illustrate the applicability of the

. . . (.

Section XI procedures to the lug / clevis type' geometry.

, 3.3 K IC Data for Lugs and Clevises - ' ~ . ..

K IC data as well as NDT data are often lacking for" supports that have been fabricated or have b'een in service. Material property testing is, therefore, necessary for any toughness evaluation, either by LEFM or NDT temperature.

For the fracture mechanics procedures to be useful', material .

1 testing of KIC must be performed on the lug and clevis materials, unless a sufficient generic data base already exists. .n.

3.4 Reference Flaw A reference flaw of 10% of the section ro - rj and extending entirely through the thick' ness in the location of th'e highest _.

stress concentration (radiating from the pin hole nonnal to the axial force) is recommended. For tile geometries of interest, this flaw ranges from .25 to .77 inches' deep and from 2.2 to 7.75 inches in the lug thickness direction.

The loading conditions for which the toughness is evaluated are the maximum Level D or faulted conditions. Normal loadings are f .

l g .

t * .

• much lower and do not cause stresses which would produce' fatigue crack growth. Any flaws which could actually be n... .~"-'

. : n.

present in these supports,therefore, must' b. , .e prestnt"*

prior to service. Cracks of the ~ magnitude' of th'e reference - '.

n .~

flaw, therefore, are clearly significantly greateY than~' "

surface flaws that would be easily detected by'preservice C '-

inspection. - ~ " ' '

Reference flaws of 10% of the'secti'on and entirely 'through the thickness would enable demonstration of an 'ade'quate' '

~'

, safety margin against brittle 'fra'cture. ~~' '

3.5 Acceptance Criteria Using the design basis loads'and the reccmended reference e

flaw size, a Ky can be computed for each geometry of'the"' '

lug / clevis type. Integrity is assured if:' '

Ky <

KIC where K IC is conservatively determined from material test data. Due to the conservatism in the loadings, in the computation of KI , and the reference flaw, no additional safety factor need be applied.

4.

_LEFM Evaluation Procedure for Succort Skirt Geometrics ' ' '

The LEFM procedure is based on a simplified conservative ~ stress ~ ~

analysis of the support skirt geometry.' Using the stress distribution in the region of highest stress around the circumference of the-~

skirt, and assuming a crack in the skirt' to flange weld, a stress -

intensity factor, K ,can I be computed. ASME B+PV code Section XI procedures are suggested for the computation 'of KI and the degree of conservatism in this process is established by a comparison '-

with a d2 tailed finite elen:nt analysis of thicracked g::ometry.

Residual stresses are addressed and an acceptance criterion based on X IC data is recomended.

4.1 Stress Analysis of a . Typical Cylindrical Skirt A typical cylindrical skirt is analyzed by traditional methods '

. in the stress report. The greatest loads around the circumference are compoted from the uplift and overturning forces on the skirt. These loads consist of an axial force and possibly a bending moment on the skirt. The stresses in the weld region (see Figure 4) can be conservatively determined simply by

~

the P/A + !!c/I bean formula. The stress concentration effect ir minimized by the presence of large fillets at the intersection of the skirt and flange.

For skirts which are furnace stress relieved after welding, -

residual stresses are minimized. The maximum possible

. residual stress, resulting from such a procedure is about 8 ksi distributed in a cosine like pattern with tension ,

on the surfaces and compression in the center of the weld.

It is most likely that the reference flaw in the weld region would be so long as to relieve the residual stress, that is, it extends beyond the tensile region, and eliminates the -

necessity for consideration of residual stresses.

4.2 K Determination g ,

The stress intensity factor, Kg, can be conservatively computed from the stress distribution in the uncracked structure by the procedures of ASME Section XI. 'An example of this process is given in Figure 13. For this case t

a crack is assumed to exist in the flange to skirt weld in the plane of the top of the flange. The crack depth

• . l is assumed to be 0.5 inches measur:d from the outside surface of the skirt. The fillet is also cracked but is neglected

-in this simplified analysis. The stress intensity factor is computed to be 0.863 Ksi / kip load / inch in Tension and 2.38 Ksi d/kipinch/inchinbending

• 4, A finite element analysis of the same geometry was performed using the MARC general purpose finite element program with a J-integral procedure for the computation of K y

. The' finite element model is shown in Figure 14. The crack tip elements represent the appropriate elastic singularity for LEFl!. The

, Kg computed from the finite element analysis is 0.509 Ksi b/ kip load / inch in tension and 0.96 Ksi h/

kip inch / inch in bending As another example, a .25 inch crack in the weld is ~ -

assumed. For this case, Kg computed by the Section XI procedure is 0.414 Ksi h/kipload/inchintension and 1.47 Ksi N / kip inch / inch in b'ending.

and the finite element analysis values are:

0.312 Ksi M/ kip load / inch in tension and 0.69 Ksi h/kipinch/inchinbending..

Variations on the skirt geometry have been evaluated by the finite element analysis and the Section XI procedure with consistant results. Comparison of the Section XI and Finite illustrate the conservatism and applicability

~

Element results of the Section XI procedures to the skirt geometry.

l

--- - ---~ - :-N :  : - ?~=~

A

~ - -

4.3 K IC Data for Skirt Materials .. a

3. . . ame Skirts are frequently fabricated from pressure boundary materials' for which considerable toughress: data already exist.c For materials where toughness data are not availableitesting  :-

must be performed. - ..

4.4 Reference Flaw .. .. . .

c q..

A reference flaw of 25% of the skirt thickness with a/ maximum flaw -

of 1 inch, and extending the full. depth of the slice. of skirt unoer consideration is recomended. This quarter thickness' flaw is consistent with ASME III Appendix G (Reference.9) analys'is frst welded vessels. In the case of the pressurizer skirt, this flaw size is 0.35 inches plus the .5 to 1. inch fillet radius.

By the same reasoning used in Section 3.4, since ' normal loads would not tend to open or extend the crack, any. crack present must exist prior to service. flomally required inspections eliminate the possibility of the existence of such a crack thereby assuring that it is conservative as a reference

~

( flaw. .

4.5 Acceptance Criteria - -. .

1 Using the design basis loads, residual stresses if required, and the recomended reference flaw size, a Kg can be computed

. for the skirt weld. Integrity is assured if: - .i.

l . .

i i Kg < K IC where K IC is conservatively determined from test data. Due to the conservatism in the. evaluation no additional safety factor need be applied.

..g. ;

v - -

4

5. LEFli Evaluation Procedures for Beam Column Sections _

The LEFM procedures for beam-columns are essentially the same as for skirt geometries for cracks assumed.near the column to baseplate interface . Section XI procedures are used ' - - -

directly for assumed crack locations remote frorc-theC- . :5 '

~

ends as shown in Figure 15. In case; loads are applied in a manner not considered in Section XI procedures (such as torsion),

Xg formulas from accepted handbooks, such as Reference 10 should be employed. .- - - -

Finite element analyses previously' performed by CE have demonstrated

~

that the Section XI procedures and handbook solutions produce -

conservative Kg solutions for ,the solid section beam-column geometry. .

A reference flaw size of 25% of the beam-column thickness up to a maximum depth of 1 inch through the entire width of the section is recommended.. For many beam-columns in typical service, no-stresses exist to cause crack propogation during normal operation, and, therefore, any flaw must exist prior to service. Tha-reference flaw recr=anded would be easily detected by normal preservice inspection.

For beam columns which are stress relieved after welding, residual stresses are mininiized and need not be considered -

in detail. Beam columns are frequently fabricated from- - -

pressure boundary materials for which considerable toughness data already exist. For materials where toughness data are not available, testing must be ' performed.

Me O q e- -

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

6. Preservico Inspection _ .

, t Typical industry practice requires non-destructive : examinations and

. inspections to provide assurance that injurious crack-like defects i do .not exist in both material' and finished product fonns. Inspection techniques include visual liquid dye' penetrant, magnetic particle,

!, ultrasonic, and radiographic. Stringent acceptance c'riteria found -

in both ASME and ASTM requirements do not permit theexistance' of i defects wh'ich are of the magnitude conservatively recommended herein j for reference flaws. The vast' majority of RCS supports are not subjected to cyclic loadings sufficient to~ cause fatigue: crack growth and therefore, inspection at the fabrication stage is a reasonable and sufficient basis to provide assurance that potentially-injurious cracks have been identified.

_f

. . a. ... ..

g 9 9

• O

.. .- J .' ..

J 4

6 .

s e- S ,

O #

ee . e 6gy e, m ee * * . . a 4

. s . -s

. ,  !.' S w ,

~ ~

. .. . , . . J. :

O e a

f ,

7. Sumary of R:;comendations

1. Fracture cechanics could be applied beneficially .

. to several support types including lugs, clevises, skirts and beams-columns. , .

.c. . %, ef m

2. Kr solutions based on classical stress analysis ,. ~

and Section XI procedures have been demonstrated to .

be conservative,yet reasonable for the geometries .

addressed,by comparison with detailed finite ~ element' ' ^' ~ ~

analyses. . , ,

3. Residual stresses are small in non-welded and stress -

relieved structures. The presence of flaws as large "

as the reference flaws recomended couid ~ ' ' '

relieve most of the residual stress in the flaw region. -

,, .,n. .. , ,.

4. KIC is available for many suppor't materials'. . . . , , - '

Toughness testing will be necessary for those ' .

materials for which XIC data are not available.. _,

5. Reference flaws of 10% of the section thickness - * ~

~'

in the highest stressed region for nonwelded geometries are recomended. For welded geometries, reference flaws of 25% of the section thickness, with a maximum flaw depth of 1 inch in the highest stressed region are recommended.

These reference flaw sizes are consistent with other cccepted reference flaw procedures.

6. The conservatism in the selection of Icad cases to be considered,the computation of stresses and Kg, the size of the reference flaw and the '

lower bound K philosophy, all combine to produce an adhuately conservative evaluation. -

. .?

e e.-- .

G

-e

L

8. References

1. Potential for Low Fracture To6ghness and lamellar .

Tearing on PWR Steam Generator and Reactor Coolant .

Pump Supports, HUREG 0577 for connent, U.S. NRC.-n .

October, 1979

2. NRC Memo, D. Eisenhut to Licensees, May 19 ,[ 1980 ,

3. HRC Memo, D. Eisenhut to App.lic_ ants, May 20, ~1980 ,

4. Flaw Evaluation for Support Structures, EPRI Project 1757-2, Presented by R. C. Cipolla, December 10, 1980 , ,

. 5. Summary of December 17, 1980 Meeting Regardirs Inclusion of Linear Elastic Fracture Mechanics in the Resolution of

~

USI A-12(Potential for Low Fracture Toughess and Lamellar Tearing on Component Supports), HRC Memo by Richard Snaider, December 30, 1980 ,

6. ASME Boiler + Pressure Vessel Code Section XI Appendix A 1980 Edition - -

7. Timoshenko, S. P.; Goodier, J. M., " Theory of Elasticity",

3rd Edition, McGraw Hill Book Co. 4970, pp. 138-139

8. MARC-CDC Non-Linear Finite Element Analysis Program,1 Control-Data Corp., Minneapolis, MN, 1976- -

9. ASME Boiler and Pressure Vessel Code Section I.II, Appendix G, 1980 Edition

10. Tada , H. , Paris , P. C. , Irwin,; G. R. ,- "The Stress Analysis of Cracks Handbook", Del .Research Corporation, llellertown Pennsylvania, 1973 ..

t

.i.

rn

'l e

4

--- e

. . _ .__ .. \

,;  ; .  ?. +

1 >

. i 4

MAX.

MIN.

r; 4.3" 2.75" -

. ro 12.0" 5.00"

l. 98.5" 21.00" -

t 6.0" 2.75" I

e t i IT V 9)

I '

/

/

t

?

ll i f>

e n

' e I

I Figure 1 TYPICAL LUG GEOf.1ETRY AND A RANGE OF KEY Dif.1ENSIONS i

D E

a

..  ?',*_+.--.,..,,,. ,,.--,.,.,.,,_,,n..,. ,n.,-._ _ , _ , , ,v_ _,g,..,,,,,,,,,.,._,,,,,.,n,,,,,e_.,,,,_,_l,,, , , _ . , . , , , . , ~ , , , , . -

, 3 _. ._. . _

-o _

MAX. , ... M I N. , -

~ '

r. 3.75" ~ 2.5" .

r, ~ ~9 00" ~ ~ ~ 5 0" t 7.75" 3.5" I

o i + t #

I

, 'i

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

i

- c . . ,

r J

Figure 2 TYPICAL CLEVISE GEOMETRY AND A RANGE C:. KEY DIMENSIONS e

GW0 -

e 9

g .

e e

0 e

., - . . - - -. y y . - . - , , - - . - - , , , , - .y.-,,, w , . . - . , , - _ . . . - . , . 9-

..,,3,, ,- . .

i t

i 4

+t +

l ro MAX. MIN' rg -

]f 6.0" i .0" l i r; ro 12.5" 5.5" l

l .t . 5.7" 3.625"

-__. _ _ -- - _\ _ ____

8 i .

- MW' l l

. l, i '

l. _) (_

8 e ,

8 l

h g

l l .

Figure 3 e ,  ;

j TYPICAL SHORT LUG AND RANGE OF KEY DIMENSIONS I

e 1

, an

~'

r-

. l

.- ,i ,

.-- -- . - t. .. . .

. e t ,

6.625 >- 13/8" "- 100.75" 1.D.

WELD REGION 1" + 1/4" (TYP)

. /s / s -

) ,

31/4" Y

87/8" z 1

Figure 4 .

SUPPORT SKIRT GEOMETRY USED ON THE PRESSURIZER G

e

-- , ..e--,.-.- g .,--,.r,--- r--- , , c- - ,- - , ---.v--er,,,.a.,.,, .--.-n,, , - . - - -. , - - - - - ,, n ,. - --- -,- - -- - . , -- -

• C -

O O.O' O F-l O!

o4-.jl

..-,.. ll 1 I I C L. _J 0 O 'O l

, r- , r

% WELD REGION

_ w- - t _ .

1. .

~ .

l'

% WELD REGION

i i i

i. l I

~ , ~q-i ,

7<s,,,,,,ee i

l TYPICAL '!

DIMENSIONS __

t = 11" 97

//

p W = 30" ---

f,!/ ---

.l> 200" . . .

l 24 ,

1 BASE PLATE Figure 5 TYPICAL SUPPORT BEAM COLUMN

.g--._ y,., , ._ ~ - ,,.r, .,,-....m.. . - , , . ,,__....__._,_..-_._-.,m,. . . , . , . - , . - .

, ~ .

A 2 i .i i 1 - <- -

O N

x -

9 1 , .

1 o -1 -

z C

en 9:

x .

tn N

• 2 .- _

= ,

CC O

u.

Di M

z -3 -

-o

\

4 - _

t -

\

5 - - -

-6 I I I I 4 6 8 10 12 r , INCHES , ,,

Figure 6 STRESS DISTRIBUTION ACROSS THE THICKNESS FOR LUG WITH r; = 4.3", or = 12.0"-

T-+ m e = g. e- -.wy -y *p-- - - , - -

es9 .-p,, -,y,,,,-,,w*,+-i- -- , y- sp9 -- in, - - ---g-y--me p p,-w-* ye-,-,g-- -

aee

. . c. , ,

• - 4 i g i i i i 3 -

2 - ,

1 .

M 0 9 - -

I I-I o -

E .1 -

M x -

n .

n ~

i .2 a

C.

C*

O u.

~ -

y .3 - -

2_.

o 4 _.

5 - , ,

6 _.

' I I I I I 7 i I 4.0 4.5 5.0 2.75 3.00' 3.5 r, INCHES Figure 7 STRESS DISTRIBUTION ACROSS HOLE TH'CKNESS FOR LUG GEOMETRY

~

r; = 2.75", ro = 5.0"

. y ., .~, -c.-- , ,- +- - g.-- --- - , - - _ - . - - - - - - . . -

1 PE ss.. . .

, v :r . . . .

.gg ,

1

.~,.,. .

. J l

/ . i

. gi. i .

,,2:  ;

y ;i.. l <

2M i -

4[

ccap~p -

mr ~

i -

?

8 n,w;;v ! 1 b I5f),~ *' '

i C l> -

e~- _ ;f

• j mr. . :. --Is a L-" 'l

- 3 ~

.e,&g

RS

) r.

4~ -

< t ~

t *~f > j

>e,*. *

,c -

,' 1 Q9 ~

r .

j -

wI.l

, L Q .

v4 A

-;q -

1 -

P, r -

r ry..

A lbQ

.ry:

J u O ,

R 1

e g -1

'f 1.0 l 1.5 riry 2.0

. STRESS OlSTRIBUTION ACRosstgg . G},,Dlt,1ENSIONAr

c. , , , , . enn..,

u - _

4 I 3 - --

2 - -

c.

2-N l - -

O O

I

-1 .

1.0 1.5 2.0 r/r; Figure 8

, STRESS DISTRIBUTION (IN NON. DIMENSIONAL FORM)

ACROSS LUG HOLE WITH ro / r; = 2 t

e

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

l 2

i i -

. 1 - ~

c.

5 N .

O 0 -

I I 1

1 2 3 4 rhi Figure 9 . .

STRESS DISTRIBUTION (IN NON Dir.1ENSIONAL FORM)

ACROSS LUG HOLE WITH gr /r;= 4 o

O e

,--------9-_p .p.--- _ .-... -- -_ . , .- ,z.,p, , ,._. ,s.,- . - - , , , .__-w_,_n., ,-7-_, . , - _ , - ,--.,_._..,,p.- , - - -- -

l ,

e

~'.

ob ACTUAL NONLINEAR STRESS DISTRIBUTION o

l 5g L l s l  %

l 1

l I (N EQUIVALENT LINEAR REPRESENTATION OF STRESS DISTRIBUTION I

l 8m '

I j++ f,+ *

• E l I a l

t 8 W R f FLAW ~7Ml -

l (a/f= 0) y t & .

I I t = WALL THICKNESS I l l /

l /

I / .

- l / h' N S'U V

K) = o mM m/a/O + Mob bEM l

WHERE

'me 8b = MEMBRANE AB)D BENDING STRESSES '

a = FLAW DEPTH .- - - . - . _ . .

O = FLAW SHAPE PARAMETER = 0.8 i

Mm = CORRECTION FACTOR FOR MEMBRANE STRESS

= 1.2 FOR a/t = 0.1 FA b = CORRECTION FACTOR FOR BENDING STRESS

= 1.03 FOR alt = 0.1 l .

l . Figure 10 l .

ATYPICAL EXAMPLE OF Kg CALCULATION FROM ASME SECTION XI 1, .

I i .

N N

/ \ _

J n, 7 ,

2 Figure 11

~

1 FINITE ELsMENT MODEL OF CRACKED LUG GEOMETRY e

e

• e e

.. _ . . - , c. , . . , . . . _ _ - _ . , , ,.,. . ... .,~. . , . , _ . . . . . , _ _ _ . .

~

at s

2 .

1 Figure 12 .

- CRACK TIP MESli FOR LUG GEOMETRY

^

.; ~

e .

FROM SECTION XI APPENDIX A: P M K=o g mMm 8 d8/O

+ abMiA ga/Q WHERE P IN KIPS 4 6.625" > 13/8 4 100.75" 1.D.

M IN KIP-INCHES

~

~

1" + 1/4'" (TYP) a TENSION CASE om= P/1.375 KSI/IN j FOR a/t = 0.5/1.375, Mm"l9 i 0.5" ~ a e Kg = P/1.375 x 1.9 x 8/0.5/1 31/4"

~

= 1.727P KSI8/IN ' y b BENDING CASE

< 87/8"  :-

ob = Mc/l = M x 3.174 KSI/IN FOR a/t = 0.5/1.375, Mb=1.2

. 111" BC Kg = M x 3.174 x 1.2 x VFV 0.5/1

= 4.761 M KSl8/IN .

SINCE SECTION XI PRESUMES SYMr.1ETRY IN THE CRACK PLANE,THE SECTION XI FORT.1ULAS RESULT IN A K VALUE DOUBLE THE VALUE FOR THE NON SYMMETRIC CASE HERE. THEREFORE, g i

Kg = 0.863P KSI/IN/IN KB3 = 2.38M KSifiM/IN i

I Figure 13 i SECTION XI ANALYSIS METHOD FOR SKIRT GEOMETRY l

i l

i

u .~ --.

a, e

• l s

N (f .

\\ '

)Q\ (- :

/ '5M \ \

. l

\ #t%j

/ /V F Figure 14 FINITE ELEMENT MODEL FOR SKIRT SUPPORT GEOMETRY

$ O

- - . , , - - - - , ., - - , . . . , , - ---,-~,,w n.-.----,.-,n,--,---,-.---,...-------------<-~.-,-w-n - - - - - - - ~ ~ ~ - - - - - - - -

v Figure 15 e IDEAllZATION OF DEAM COLUMN GEOMETRY g FOR K EVALUATION O O.O

" O p- g

• Q l; j i ' ,

- i I I

' ' r-I',*l -t-e ,e g ..

I 3

I l' C L_ . .J 0 O 'O [

SUBSTRUCTURES FOR

' Ki EVALUATION BY PROCEDURES OF V

SECTION XI.

REFERENCE FLAW ASSUMED IN V! ELD .

,< ) <

I ,

--i

!  ! y  !

nn , sn j/ s ,,,,gg

/

SUBSTRUCTURE FOR Ki EVALUATION

./ -

BY PROCEDURES OF SECTION 4. REFERENCE t

FLAW ASSUMED IN FILLET REGION.. _

\

i

/p/-.

/,

i

/ _

i i l

BASE PLATE l

t

1. e E -

a e

ATTACHMENT 3

- .,