ML18037A269

Supplements 800415 Interim Deficiency Rept Re Biological Shield Wall.Forwards Agenda & Presentation from 800610 Meeting & Discussion Re Frame Mechanics Evaluation.Sample Fracture Mechanic Calculations Encl

ML18037A269
Person / Time
Site:	Nine Mile Point
Issue date:	06/18/1980
From:	Rhode G NIAGARA MOHAWK POWER CORP.
To:
References
NUDOCS 8006240406
Download: ML18037A269 (90)

Text

REGULATOR NFORMATION DISTRIBUTION S EM (BIDS)

P'+(0 ACCESSION NBR:8006240406 DOC.DATE: 80/06/18 NOTARIZED: NO DOCKET FACIL:50 410 Nine Mile Point Nuclear Stationi Unit 2q Niagara Moha 05000410 AUTHBNAME AUTHOR AFFILIATION RHODEPG,K. Niagara Mohawk Power Corp, REC IP ~ NAME RECIPIENT AFFILIATION

SUBJECT:

Supplements 800415 interim deficiency rept re biological shield wall. Forwards agenda 8 presentation from 800610 meeting 8 discussion re frame mechanics evaluation. Sample fracture mechanic calculations encl.

DISTRIBUTION CODE: BOI9S COPIES RECEIVED:LTR g ENCL TITLK: Construction Deficiency Report (10CFR50 ~ 55K)

SIZEI gg N OT PQ ES0~~

ES RECIPIENT COPIES RECIPIENT COPIES ID CODE/NAME LTTR ENCL ID CODE/NAME LTTR KNCL ACTION: AD C (gg Ngj'AIC-21 1 1 BC gd uhlagCOOgi 5 1 1 LARuSeBROak; 20K. 1 1 PMklp8g g 1l 1 1 INTERNAL; AD/RCI/IE 1 1 AEOD 1 1 EDO 8 STAFF 07 1 1 EQUIP QUAL BR16 1 1 FIELD COOR/IE13 1 1 HANAUERgS ~ 1 1 HYD/GEO BR 22 1 1 ISE 05 2 2 LIC QUAL BR 1 1 MOOREPV ~ 1 1 MPA 08 1 1 NRC PDR 02 1 1 OELD 23 1 1 OR ASSESS BR 17 1

~~l PROC/TST REV BR

'-'XTERNAL:

1 1 1

QA BR STANDRDS DEV 11 10 1

1 1

1 ACRS 24 16 16 LPDR 03 1 NSIC 04 1 1 JOI 25 ]g8O TOTAL NUMBER OF COPIES REQUIRED; LTTR 41 ENCL 41

I rg ff P

i l i c 'I i L

I A

Wi ~

E

NIAGARA MOHAWK POWER CORPORATION/300 ERIE BOULEVARD WEST, SYRACUSE, N.Y. 13202/TELEPHONE (315) 474-1511 June 18, 1980 Office of Inspection and Enforcement Region I Attention: Mr. R. T. Car lson, Chief Reactor Construction and Engineering Support Branch U. S. Nuclear Regulatory Comission 631 Park Avenue King of Prussia, PA 19406

Dear Mr. Carlson:

Re: Nine Mile Point Unit 2 Docket No. 50-410 This letter supplements our April 15, 1980 10CFR 50.55(e) interim report on the Nine Mile Point Unit 2 biological shield wall. The attached agenda and presentation from the June 10, 1980 meeting with the staff of the Office of Nuclear Reactor Regulation regarding the Nine Mile Point Unit 2 biological shield wall deficiency are submitted for your information. This presentation is the same 'as the one made at your offices, on May 7, 1980, except for the addition of the discussion regarding the fracture mechanics evaluation. The presentation made by our Architect/Engineer (Stone 8 Webster) regarding its Procurement guality Assurance Program was also modified somewhat from the May 7, 1980 meeting.

Also enclosed are sample fracture mechanic calculations which show the methodology used in the analysis. A typical set of calculations will be available for review by your staff at our Architect/Engineer offices after July, 14 1980.

I l

V

It is Niagara Mohawk's understanding that the review and approval of our method of closure of this item is the responsibility of the Nuclear Regulatory Commission's Office of Inspection and Enforcement. The biological shield wall deficiency does not involve a, deviation to the Preliminary Safety Analysis Report for Unit 2.

Very truly yours, NIAGARA MOHAWK POWER CORPORATION G ral . Rho e Vice President System Project Management PEF gk Attachments xc: Director of Inspection and Enforcement U. S. Nuclear Regulatory Commission Washington, D. C. 20555 Director Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D. C. 20555

=~

l.

I

AGENDA Nine Mile Point Unit 2 Biological Shield Wall Meeting, 'June '] p '] 98p NRC'ethesda, Md..

Stone & Webster NMPC .

Attendees Attendees L. B. Hirst S. F. Manno C. E. Crocker C. D. Terry D. A. Boe K. Ward S. A. Ali P. Francisco I. Sprung R. G. Burns R. Patch R. H. Pinney F.. Feng C. F. Reeves W.. Di,ehl'...

V'. Zilberstein Introduction A. Purpose of Meeting NMPC/C. D. Terry B.::S'tatement of Problem NMPC/C. D. Terry C. Overview of Presentation NMPC/C. D. Terry Biological Shield Wall Description A. Functional Requirements and Physical Description S&W/S. Ali B. Design Criteria S&W/S, Ali C. Analytical Techniques.. S&W/S. Ali D. Method of Fabrication S&W/S. Ali E. Spec. Requirements S&W/S. Ali F. QA Program S&W/R. Burns III History of Events NMPC/K. 'Ward IV Metallurgical Discussion S&W/D. Boe V Closure Plan S&W/C. E. Crocker VI Engineering Evaluation A. Stress S&W/C. E. Crocker B. Fracture Mechanics S&W/j. Sprung C- Summary, of Evaluation Details ., S&W/C. E. Crocker VII Summary NMPC/C. D. Terry

~ 'I a l'

II BIOLOGICAL SHIELD WALL DESCRIPTION A. Functional Re uirements and Ph sical Descri tion The biological shield wall is a composite steel and plain high density concrete, cylindrical shell placed around the reactor pressure vessel.(Fig.l)

The shield wall's main function is to provide radiation- shielding and accommodate pipe restraint loads. It also provides anchorage support for floor beams, pipe supports and insulation and provides support for the star truss/stabilizer system.

General features of the biological shield wall are as follows:

a. The shield wall is approximately 20 'inches thick, 28 foot 2 inch I.D. and 48 foot 4 inch in height. (Fig. 2)

b. The inside and outside steel shells are constructed of 14 inch

'plates stiffened by horizontal and vertical ribs, also 14 inch plate A537 Class 1 (Gr. 50) steel is used for steel shells and stiffeners.

c. The 1 foot 5~< inch plain concrete fill in between the two shells is "nonstructural" and is provided only to satisfy the shielding requirements.

d. The shield wall is anchored to the concrete reactor support pedestal at the base, and the top of the shield wall is supported fiom the containment wall by the star truss assembly. The support at the base allows for radial growth while restraining rotation and tan-gential movement.

e. Major pipe penetrations are sealed by steel doors which are designed to:

1. Provide the required radiation shielding.

2. Resist transient pressure loadings within the shield annulus.

B.'esi n Criteria The BSW is designed in accordance with the AISC code for the normal operating load conditions. For the abnormal/extreme environmental load combinations, the allowable stresses are increased in accordance with the factors specified in the NMPZ PSAR.

The following major loads have been considered in the analysis and design of the BSW:

Deadload and seismic loads Accident temperature cases consisting of the maximum temperature differentials between the inner and outer walls occurring as the result of a loss-of-coolant accident (LOCA).

1 0 'x

Accident pressure differential between the inner and outer walls occurring as the result of a LOCA.

- Pipe restraint loads occurring as the result of restraining pipes following a rupture.

Jet impingement loads resulting when pressurized fluid from a ruptured pipe strikes the BSW.

C. Anal tical Techni ues The structural analysis of the biological shield wall is performed by the finite element method using the computer program STRUDL. The shield wall is modeled using a 180o model with the appropriate boundary conditions for the symmetric and anti-symmetric loads. Analysis for general loading conditions are conducted using principles of superposition.

The inner and outer walls as well as the horizontal and vertical stiffeners are modeled using isoparametric elements. The stresses in the biological shield wall under normal operating conditions are very low. The conditions which control the design of the biological shield wall and under which the stresses approach the allowables are the accident conditions, such as accident temperature and pipe rupture loads. The stresses from this analysis are used in the engineering analysis to be discussed later.

D. Method of Fabrication The biological shield wall is divided into three rings as shown in Fig. 4.

The three rings are further subdivided into three 120o segments, a total of nine segments. These nine segments were fabricated in the shop and then shipped to the NMP2 site. These 120o segments were welded together along vertical seams to form the three complete rings at the site. The three rings will in turn be welded together along the horizontal seams to form the complete biological shield wall.

E. General S ecification Re uirements All work was performed under seller's QA program Welding was done in accordance with AWS Dl.l The following options were provided to perform NDE of all full penetration welds:

1. Radiographics or ultrasonic inspection with MT of root pass

2. Progressive magnetic particle at 1/3, 2/3 and 3/3 weld joint thickness The seller chose to employ "T along with MT of the root-pass

- AISC code of standard practice was invoked for worhnanship requirements.

~ r ~ 1 t p, 4

~ 'L%

REACTOR BiLI%.DING CONFIGURATION i

REACTOR P' REACTOR PRESSVRE VESSEL (R.P.V.), .

EL. 316'-1 '/z STAR TRUSS 28'-1'/z" i EL. 302'-0 TOP OF BSW 1'-8'/g" BEND LINE EL. 314'-1~5" 67'"

.PRIhhARY BIOLOGI CAL CQ NTAINMENT Rf ACTOR SHIELDNIALL

.SVPPORT SKIRT EL. 266'-'6i/z" 1St 31 ~

DRYWELL EL 240'-0" BEND LINE R EACTOR PEDESTAL I

I 9 1 0'li

~

EL. 175'-0 MAT FIGURE I

~ Af VERTICAL STIFFENER 1'-8.1/2", R 1 1/2" (TYP.)

1 1/2" (TYP.)

FIGURE 4

~ ~

E

BASE DETAllL 15'-9 3/4" RAD 14'-0 3/4" RAD l'-5 >/2" 8 3/4" 11/2" 1 in" INNER SHELL PLATE

c.'p, ~": 4 3/iI 1/2 OUTER SHELL PLATE GROUT AREA VfASHER PLATE

PASE PLAT Qll

~ .o'5:c

'. ~ ~ ',

QDt go:<

O..g EL. 265'-5 >/2" 3ll

~ ~

'. PEDESTAL SOLE PLATE 0, ANCHOR BOLT FIGURE 3,

3RD RING 17'-6 1/2>>

INNER SHELL 2ND RING PLATE 15'-1 5/16>>

14'-0 3/4>> R CONCRETE OUTER matt COVER PLATE 1ST RING HORIZONTAL 15' 3/>6>> STIFFENER BASE PLATE STAL FIGURE Q

I F.

The QA Program imposed by the specification requirements resulted in certain actions by both S&W and Cives (the Contractor).

Specification No. NNP2-S204G was reviewed and approved by S&W's QA Department (Quality Systems Division) to ensure the inclusion of appropriate QA/QC requirements.

The specification and associated shop test, inspection, and documentation (TID) report required, in summary, the following:

Cives Compliance to Appendix B, 10CFR50 Submittal of QA program Transmission of QA requirements to identified subcontractors 4, Conformance of NDT to AWS Dl. 1 S. Welding and NDT procedures submittal.

S&W l. Qualify Gives 'as Seller by survey and audi.t 2t Perform inspection (over a 26-month period) covering 44 specific TID attributes Of the 44 attributes, 11 specifically address weld quality (see attributes applied) ~

Attributes Applied The following veld-related attributes were applied:

Welding Procedure Electrode Control Procedure Qualification of Welders Weld Preparation Weld Inspection Random Check of Fabrication Completeness Inspection of Surface Defects NDT Test Oper Certifications Procedure for NDT Inspection NDT Inspection of Welds Reports of NDT Tests The above comprised a normal QA/PQA effort on this type of structure and we believe is consistent with a properly implemented activity.

QNQC ACYIOHS NIIME MILE 2 BIIGLGGIICALSHIIELD WAI L

e REVIIEW 8 APPRGVAL GIF CIVES IMDT PRGCEDURES o FABRIICATGR SURVEY 8 QUAILIIIFIICATIIGM

CN'ES PREVEMY)VE MEASURES e CGMPLIIAMCE TG JOCFRGO, APPEHDIIX 8 PASS AILGHG GIF QA REQUIIREMEMTS TG SUBS e MDTTGAWS 01.)

HOT GPERATGRS TG MEET ASMT-TC-1A

e YEST IIK~SPECY~GH, QGCUMEHYAY(GN-44 AYYBIIBUYES 0 II I - AYYBIIBUYESAQQBESS NELQIIFJG 9 3- IIHSPECYGIRS, 26- MGMYHS, 4,SOO-MAM-HGURS

e PIRGBLEMS IIDEHTIFIIED %EIRE IRESPGNDED TG e PIRGGIRAM MEASUIRES ADEQUATE

III. HISTORY OF EVENTS INITXAL PROBLEM WAS FIRST DISCOVERED AT TOE OF THE OUTER COVER PLATE TO BASE PLATE WELD DURING FIELD INSPECTION.

S&PLE OF WELD WAS SELECTED FOR UT TO VERIFY WELD WAS FREE OF DEFECTS.

WHILE NO SIMILAR DEFECTS WERE DETECTED IN TOE AREA, INDICATIONS WERE FOUND XN THE ROOT AREA OF THE WELD WHICH PROMPTED l00% INSPECTIONS AT THIS TIME THIS PROBLEM WAS REPORTED TO THE NRC AS A POTENTIAL 10CFR50.55(e).

A SPECIAL TEST BLOCK WAS DEVELOPED SUCH THAT ANY REFLECTORS FROM THE BACKING BAR COULD BE PROPERLY IDENTIFIED.

APPROXIMATELY 20% OF THE LENGTH OF THE WELD REQUIRED REPAIR.

METALLURGICAL SAMPLE WAS REMOVED FROM OUTER COVER PLATE TO BASE PLATE WELD FOR ANALYSIS'

15'-9 1/455 RAD 14'-0 3/4 RA D 1'-5 >/255 8 3/455 1 1/2." 1 >/255 INNER WALL PLATE ':.o.

u.-..() fi' @~4:o" 55 c.P 4 3/S'UTER 1 1/2 NALL PLATE GROUT AREA WASHER PLATE

) 'PASE PLAT) qr)C oD0 95)

'> o'Po o

~

~

oQc 0 ~

~

Q.

'>C'5$

Q(

Oo<

~ +c. - EL. 265'-5 t/255 o4 255 f55 SOLE PLATE a*.

~ ~o;.,

'~0 PEDESTAL ANl" HQR DOLE FIGURE 2

CGVER PLATE TG BASE PLATE 57t"-LD TYPICAL 5/8 11/Z" 13/16 45 COVER PLATE 1/4 45o TACK WELD 1/4 X 1'ACIilNG BAR BASE PLATE NASHER PLATE FIGURE 5

VISUAL LINEAR INDICATIONS WERE FOUND IN HORIZONTAL STIFFENER TO INNER WALL WELDS DURING ONSITE INSPECTION OF THIRD RING SEGMENTS.

INVESTIGATION CONDUCTED OF ROOT AREA BY MAGNETIC PARTICLE TESTING AND GRINDING ON ALL THREE RINGS-MT EXAMINATION YIELDED REJECT RATE OF APPROXIMATELY 22% ON RING 3 SEGMENTS. RINGS 1 AltD 2 WERE ACCEPTABLE.

HIGH RATE OF REJECTION OF THESE WELDS LEAD TO DECISION TO CONDUCT AN INVESTIGATION OF ALL WELD CONFIGURATIONS.

I l

HGPIZGNTAL STIFFENER YG INSIDE AND Gum SIDE MfALL PLATES 1/2" I

0 3/16 8

30~

I TACK TACK INSIDE V/ALL WELD PLATE 24'OOSE HORIZONTAL OUTSIDE STIFFENER WALL COYER PLATE FIGURE 8

0 SINCE MAJORITY OF MELDS MERE INACCESSIBLE TO A NORlfAL AMS Dl.l UT INSPECTION, A SPZPLING PLAN WAS BPLOYED TO ASSESS MELD QUALITY.

ADDITXONALLYFOUR SAMPLES MERE TAKEN OF HORXZONTAL STIFFENER TO INNER MALI, MELDS FRO>f RING 3.

RESULTS OF SPILE PLAN DXD NOT PROVIDE DESIRED CONFXDENCE LEVEL FOR QUALITY OF MELDS.

General Weld ualit The welds that we have been discussing were made in the shop using the fluxed core process with gas shielding, or SHAM process. All welds examined 'metal-lurgically exhibited overall high quality. No cracking has been found in weld metal; all cracking has occurred in the HAZ, or else outside the HAZ in the base metal.

U CLOSURE PLAN The resolution of the shield wall weld defects is outlined as follows:

l. A 100% UT inspection is being performed for accessible shop welds (.90%

of all shop welds are accessible for UT) . A standard AMS Dl.l UT is performed. where access permits. Otherwise, a special UT in accordance with AMS Dl.l is performed.

k

2. Based on UT data, an engineering evaluation is performed to determine repairs required. The evaluation employs stress analysis and fracture mechanics to demonstrate acceptability of inconsequential defects. All PSAR commitments regarding loads, allowable stresses, and other technical requirements will be maintained. The UT and evaluation is in accordance with AWS Dial, Section 3 ' '

As discussed, a metallurgical evaluation was .performed on five specimens xemoved from the shield wall. Results of the evaluation are factored into our overall approach.

This program of UT, engineering evaluation, and metallurgical examination, all in accordance with AMS Dl.l and our PSAR commitments will ensure that the biological shield wall will be more than adequate to perform its intended design function.

The overall erection sequence to implement the approach outlined above is as follows:

1. Repair as required based on engineering evaluation

2. Attach cover plates

3. Pit-up and weld- the three rings together

4. Perform final PNHT

5. Set BSM on pedestal without concrete

6. Complete remaining repairs, if any

7. Attach remainingcover plates and fillwith concrete.

(Refer to Pragnet, Pig. 5)

aaaj fg'e f

I $ <,':.

~

a

~

g

= ~ 11 ca

)

4L. a f/ f l" 1, t a '.>C j/ / "I"..:I 5 r

a-=

':I:j3g.g $ :; i;i-;3j:.f~):q~,

PP cled la

/ i-"f-~ i'*

Ia i -f v aaa "Jf~r' 4 aa

~

'a

~saa

'-~ 41'l w aaa'A'AW a

<,, t't"i'l.

K. "AA ++7,

'A '>> -"

A~A'A v

tgq~ "j~ij)]

)g.);;; jg,' )')',*

'3l".(s.

') Yiga)

$ )L';.-

~" t t '7(

J t

A ly)<%Jjttf) ~

"4~4( )a l ~ g%P

,a -a ~

~ I'It aaaAa a a~~

.I,'r 1; I.'I.'I.i!f.Bf lg~ pr ~

....--,Q-e 7i g

~" ~.~

..,.;U:fl.. "Ala>>': ~ aa' AV

/%

aa j()() .v --

~

aa a ~

'aAAAAAaaAVA'a a.4 aa a AaAakie&f 3l ]~(a>> ><g )4otI').f~'f">

r~we~4&,vsa tapdLA~~a~M)+

~(3f Jg')/ff

~

<<aaa a a.,~

l ~

a a

%,A a ~$ p J ag~aajk+af Egal'lkaa:-L~AAAAa oII VC

-gg 4 ~

~a ,

r',i., (....)qM r.

. r a

,C aa

'~efi) .'

]f 0 $ t 1 ~

~ A

~COaakZk

VI. ENGINEERING EVALUATION A. Stress Analysis Based on UT data, weld defect sizes, locations, and orientations are obtained for evaluation. Stresses in the vicinity of the defects are evaluated as a result of reduced weld area due to the defects. As mentioned, all PSAR commitments regarding allowable stresses are being maintained. Once it is shown that the stresses are at an acceptable level, the stress and defect information are evaluated for fracture mechanics considerations.

B. Fracture Hechanics Evaluation In the industry today, a conservative fracture mechanics analysis is now a regular means of assuring the integrity of welded structures which real-istically contain some discontinuities in the material, such as slag, porosity, lack of fusion, etc.

Such an analysis provides a sound basis for establishing acceptability criteria for the discontinuities and thus can eliminate unnecessary repairs.

11ost structural steels when loaded under normal design conditions axe ductile and the methods of linear elastic fracture mechanics (LEAM) are not applicable. Therefore, yielding fracture mechanics is used to calculate critical conditions (critical crack size and critical stress). In this analysis we use both LERf and a technique by Dowling and Townley (Ref. 1).

This method, known as Two-Criteria Approach, has been used by others (Ref.

2-7), and covers the spectrum of conditions from brittle fracture (where LEFM is applicable) to completely plastic failure (where some form of limit analysis has to be used). The structure is built of 537 Class 1 steel, tehich has excellent fracture toughness in the longitudinal direction. In cases where it is required, the directional properties such as in the through-thickness direction are considered.

i For the purpose of this analysis, the variety of discontinuities which might be encountered in a welded structure can be reduced to two major types, namely surface and subsurface defects. The term "defect" is used here only for convenience and does not mean that a discontinuity under consideration is not acceptable.

It is assumed throughout the analysis that the applied stress is a tensile stress perpendicular to the defect, the applied loads are dynamic and there is no cyclic loading which could initiate fatigue cracking. In this analysis, surface and subsuxface defects are defined as in ASHE XI, Div. 1 (see Fig. 1).

In the case of a surface defect at the root of a weld, the stress intensity factor is given in Fig. 2.

The fracture criterion is given in Fig, 3. To be conservative, the dynamic fracture toughness KId, is used in this analysis. KId can be found from one of the known correlations between KId and the Charpy impact energy, Cv.

4 4

Also, we can plot the applied stress versus critical defect size and compare a reported defect size with the one on the plot (see Fig. 4).

For subsurface cracks the stress intensity factor is defined as in Fig. 5.

The above approach is essentially a LE&i method with some modifications.

When this method is applied to non-brittle materials, it may give unrealistic results.

As mentioned before, the two-criteria method avoids this problem, The fracture (or critical) stress is calculated from the expression given in Fig, 6.

In conclusion, we would like to emphasize once again the fact that the fracture mechanics analysis is very conservative. We use conservative assumptions for the defect size and stresses plus we assume that all the defects are sharp, which for most of them is not the case. Thus, we believe the analysis provides additional assurance for the safety of* the BSW structure.

C. Summary of Evaluation Details For the evaluation, the welds are grouped into three configurations:

inner wall to stiffener, cover plate to stiffener, and stiffener to stiffener. F'rom a structural point of view, the most important welds are the inner wall to stiffener and the cover plate to stiffener welds. Details for each configuration are as follows:

l. Inner Wall to Stiffener A special UT in accordance with AWS Dl.l is performed for the entire inner wall.'efect sizes and locations are mapped, stresses are evaluated, and a fracture mechanics evaluation is performed. Based on the evaluation accept/repair decisions are made.

See Table 1

2. Cover Plate to Stiffener standard AWS Dl.l UT is performed for all cover plates. UT examination cannot reasonably provide conclusive information required to perform an engineering evaluation for defects located in the top 1" of the weld near the nose of the stiffener. Therefore, defects located in this area are repaired to a depth of 1". For defects located Xn other areas, a special UT in accordance with AWS Dl.l is performed, defect sizes and locations are mapped, stiesses are evaluated, and a fracture mechanics evaluation is performed. Accept/repair decisions are made based on the evaluation.

'See Table 2

0

3. Stiffener to Stiffener All accessible stiffener to stiffener we3.ds are examined using standard AWS Dl.l where pos ible and a special UT in accordance with AWS Dl.l otherwise. Over 40% of all stiffener to stiffener welds are accessible for UT). Data from the stiffener to stiffener examination and inner wall to stiffener examination are used to predict the incidence of defect occurrence and effective loss of weld area. A special analysis is performed for each inaccessible stiffener, postulating that large amounts of weld area are removed for analysis purposes. With conservative assumptions is will be shown that adequate structural integrity exists. It is expected that all cases will be acceptable. However, in the case that structural integrity can not be demonstrated, the design will be revised to compensate. 'I See Table 3.

TABLE 1 Inner >lail to Stiffener UT Status 6/5/80 0/ No. of , Total Length  % Indications

~Rin ~Com 1. Indications Indications (in) 94 ll 3/4 97 30 1/8

.3'.8 95 17 301 3/4 TABLE 2 Cover Plate to Stiffener UT Status 6/5/80 Note: Status covers initial standard AHS Dl.l UT only ey Length  % Length

~Rin ~Com 1. ~Re '. (in) ~Re 1 100 3 j 322 17 2 100 3,838 22 3 61 824 12 TABLE 3 Stiffener to Stiffene'r UT Status 6/5/80 Standard AHS Dl.1 S ecial UT Length  % Length my No. of

~Rin ~Com 1. ~Re' (in) ~Re ~Com 1. Indications 0 0 0 0 73 54 3/8 2.5 28 0

REFERENCES

1. A. R. Dowling and C. H. A. Townley, "The Effect of Defects on Structural Failure: A Two-Criteria Approach", The Xnternational Journal of Pressure Vessels and Piping, 3, 2,77-107, 1975.

2. G. G. Chell, "A Combined Linear Elastic and Post-Yield Fracture Mechanics Theory and Its. Engineering Applications," Fracture Mechanics in Engineering Practice Editor P. Stanley. Applied Science, London.

3. R. P. Harrison, "A Unified Approach to Failure Assessment of Engineering Structures," Same as 2.

4. A. Muscati and C. E. Turner, "Post-Yield Fracture Behavior of Shallow-Notched Alloy Steel Bars in Three Point Bending," Same as 2.

5. C.H.A. Townley, "The Integrity of Cracked Structures Under Thermal Loading,"

Same as 2.

6, R. L. Roche, "Analysis of Structures Containing Cracks Some Comments on the Jl Integral Criterion," The Xnternational Journal of Pressure Vessels and Piping, 7, 1979, P. 65.

7. J. M. Bloom, "Prediction of Ductile Tearing of Compact Fracture Specimens Using the R-6 Failure Assessment Diagram," The Xnternational Journal of Pressure Vessels and Piping, 8 (1980) 215-231.

I l

k 1

SuaSuRFACE OEFEn (S ~aj 8

WHEN 8< ci, lY lS CONSIDERED AS A'SURFACE DEFECT.

1 FlGURE 1

k S

%HERE 0 a IS THE DEFECT SIZE IS THE STRESS FS = 1.12 IS THE SURFACE-ENERGY FACTOR

. FE IS THE SHAPE FACTOR FIGURE 2

4f

'I

<p

~ ~ r. ~

/

IS THE FRACTURE TOUGHNE$ $

FIGURE 3

/

~ I r

~ ~

FIGURE 0

'l II

~ ll

F I CURE 5 y ltt WHEP.E 5 = FRACTURF STRESS

= ULT)MATE STRESS 5 = STRESS CALCULATED FRGM EQUATlGNS S!bhlLAR TG THE LEFT RELATIGNS FlGURE 6

)

FRACTURE HECT lANICS - SAMPLE CALCULATIONS To illustrate the approach outlined in the report, two typical examples are given below. The first example evaluates the effect of the exclusion of one-eighth inch from the stiffener plate thickness. The second example is an evaluation of an indication discovered in the inner wall of the Ring 1.

A. 1 Surface Defect at the Root of a l<eld The exclusion of one-eighth inch from the stiffener plate thickness in the stress analysis which essentially postulates the existence of an infinitely long one-eighth inch surface defect at the root of a backing bar weld required that a fracture mechanics evaluation of such a defect be performed. The location of the, defect is shown in Figure Al.

A. l. 1 LEFM Approach The stress intensity factor for this case is given in Section YI, Figure 2 of the Fracture Mechanics presentation. The parameters in the equations are given below:

a. The surface-energy factor Fs = const = l."12;

b. The shape factor -1 FE =~-

is defined in ASME XI. This factor can be represented-as:.-

1

=

FE 1 + 4 593(a)1.65 212( )2 1 Sys where 1 is defect length and Sys is the yield strength of the material (Sys = 50 ksi).

c. Factor F(a/an) describes the stress distribution in the backing bar:

F(a ) 2 [a (a ) - 1 + -" - sin 1 ~a

]

aN 7f aN aN where a and a are shown in Figure Al.

d. Factor FG accounts for the stress field gradient caused by changes in geometry of the stressed structure. In this case the factor applies to defects emanating from the root and the toe of T-welds. Analyses of such joints were performed in References Al and A2. The most conservative of the values from these papers was used in this calculation. In this example, FG = 1.328.

1 of 8

e

e. Factor Fw is the correction for finite thickness, t, of a plate:

F w

=

2t Can-ma ma 2t

f. As shown in Figure Al, the defect size a= ha+ aN, where ha is the actual defect size and aN is the stressed portion of the backing bar. For this particular case aN = .09". Thus, in this example a = .125 + .09 = .215".

g. The stress applied + Sres where S>>pl;ed is the applied stress and Sres is the residual stress in the structure after postweld heat treatment (PWHT). The maximum design stress is used in this calculation as the applied stress so that Sapplied 25 ksi. The residual stress after PWHT at 1100 F >s assumed to be 10Ã of the yield strength, that is Sres =

1Sys = 1 x 50 = 5 ksi.

Finally, note that the defect is assumed to be infinitely long so that a/1 = 0.

Thus 2t tan-ya ma 2t 1'12 FGF+ ) 5 ~ma KI N 1 + 4 593(a)1 .65 1 'ys 212(S

)2 2 x 1 5t .2 3T

=

.215s 2 x 1.5 1.12 x 1.328 x .86

+ 4 593(0)1 . 65 21 2(25) 50

= 33.3 ksi /in.

The fracture toughness of the material is calculated from the Sailors-Corten relation (Reference A3):

K~d = 15.873 (Cv).

where Cv is the Charpy V-notch impact energy.

Although the postulated defect is apparently located in the HA7 where toughness in the through-thickness direction, according to Stone and Webster experimental data, is better than in the base metal, the base metal toughness is used here for the sake of conservatism. Based on the experimental Charpy impact tests performed at 90F and available published data (Reference A4), the Charpy energy at 100 F is:

C'> 20 ft. lb.

2of8

I Using the minimum value of Cv the fracture toughness in the through-thickness direction is calculated:

KId = 15.873(20).375 = 48.8 ksi in.

Thus, KI = 33.3 < KId = 48.8 The critical defect size evaluated from the LEFN equation on Figure 2 Aacr 4/

A.1.2 Two-Criteria Approach The critical applied stress is given by the equation in Section VI, Figure 6 of the Fracture Nechanics presentation. The parameters in

~

this equation are:

a. The ultimate stress S = (1 a/-t) =

S s'%s (1 a/-t)

St1

b. The critical stress, Sk, is calculated from the following LEFN relation:

1.65 KId 1 + 4.593( -)

1 Sk 2tL1.12 FG F( )] tan cr N

2t The above relation can be derived from equation shown in Section YI, .Figure 2 of the Fracture Mechanics presentation, except for the plastic zone correction which is not included here.

The critical defect size evaluated from the equation in Section VI, Figure 6 of the Fracture Nechanics presentation is:

~acr = .42" Using the lower of the two calculated.values of sacr,

~acr = .42" we find that the ratio cr = .42 3. 36.

aa .125 Therefore, a one-eighth inch defect at the root of a weld subjected to 25 ksi stress is acceptable.

3of8

0 I A.2 0

Subsurface Defect in the Inner Mall of the Bios iield Wall Another example illustrates an actual case of a reported indication in the inner wall of Ring 1. The indication was interpreted as a subsurface defect parallel to the surface of the inner wall (see Figure A2).

A.2.1 LEFN Approach The stress intensity factor for this case is given in Section VI, Figure 5 of the Fracture Nechanics presentation. The parameters in this equation are:

a~ Factor Nm is the correction factor for membrane

\

stress. It is given in Figure A-3300-2 of /ESNE XI, Division 1, as a function of plate thickness and defect location. For this case, i.e. defect parallel to the plate surface, Nm = l.

b. Factor FE is the same as in the first example (see Al. 1).

c~ The stress, once again, includes both applied and residual stresses. The applied stress in this example is 16 ksi, and the residual stress, as in Al. 1, is 5 ksi. Thus, S = Sapplied + Sres = 16 + 5 = 21 ksi.

d. The defect size, a, for a subsurface defect, equals half the reported defect size. In other words, 2a =

1/8 and a = 1/16 in. Note that for conservatism, if the defect size is reported as less than 1/8 in. as it is here, it is assumed that 2a = 1/8 in.

F F

e. The defect length, according to the UT report,'1 = 3/4 in.

Hence, a/1 = .083.

Thus, K 21 /m/16 I 9.13 ksi <in.

1 + 4.593(.083)1.65 -

.212( )21 50 Since the fracture toughness of the material in the through-thickness direction at the minimum temperature when the stress might develop is:

KId = 48.8 ksi Din (see Al. 1). KI <'KId 4of8

4. 0 h ~

fhe critical defect size evaluate d from LEFM equation in Section VI, Figure 5 of the Fracture Mechanics presentation, assuming that the ratio a/'I is constant, acr = 1/8 in.

and 2acr = 3 1/2 in. > t = 1 1/2 in. (plate thickness).

Two Criteria Approach The critical applied stress is given by the equation in Section VI, Figure 6 of the Fracture Mechanics pre'sentation. The parameters in this equation are:

a. The ultimate stress:

'U =

'now" 2 2 Since the defect is parallel to the plate surface, a/t vanishes, and

~s',

2

b. The critical LEFH stress, Sk, is calculated as fol lows:

a 1.65 1 + 4.5g3(-)

1 S

Id H

m na cr + .212(~)

The critical defect size evaluated from the equation in Figure 6 is:

=

acr so that 2acr = 3.4" > t = 1-1/2" (plate thickness).

Therefore, the one-eighth inch defect subjected to 16 ksi stress is acceptable.

5of8

4

REFERENCES:

Al S. Usami et al. Trans. of the Japan Melding Society, April, 1978.

A2 T.R. Guerney, "Finite element analysis of some joints >lith the weld transverse to the direction of stress", Melding Institute Research Report, E/62/75, triarch, 1975.

A3 R.H. Sailors and H.T. Corten, ASTH STP513, p. 164.

A4 J.S. Lentz, Journal of Pressure Vessel Technology, February, 1978, Vol.

100, p.77.

4 4,

Page 7 af 8

1 9 ~

4 ~

c ~, ~ < ~

4

~ P

~ 'I FlGURE A2 Pa8e 8 of 8

~ ~ g 4 ~ ~

4 I

E