ML13316B741

Significance of Circumferential Weld Indications in 24-Inch Main Steam Piping at San Onofre Nuclear Generating Station Unit 1 - Fracture Analysis

ML13316B741
Person / Time
Site:	San Onofre
Issue date:	07/31/1980
From:	Cipolla R, Egan G, Grover J APTECH ENGINEERING SERVICES
To:
Shared Package
ML13316B739	List: ML13316B741
References
AES-80-06-30, AES-80-6-30, NUDOCS 8007280404
Download: ML13316B741 (64)

Text

AES-80-06-30 APTECH engineering iervices, InC ENGINEERING CONSULTANTS 795 SAN ANTONIO ROAD. PALO ALTO

• CALIFORNIA 94303 (415) 858

• 2863 THE SIGNIFICANCE OF CIRCUMFERENTIAL WELD INDICATIONS IN 24-INCH MAIN STEAM PIPING AT THE SAN ONOFRE NUCLEAR GENERATING STATION UNIT 1 - FRACTURE ANALYSIS Prepared by R. C. Cipolla J. L. Grover G. R. Egan J. D. Byron Aptech Engineering Services 795 San Antonio Road Palo Alto, California 94303 Prepared for Bechtel Power.Corporation 12400 East Imperial Highway Norwalk, California 90650 July 1980 Services in Mechanical and Metallurgical Engineering, Welding, Corrosion, Fracture Mechanics, Stress Analysis 8007280 7 &

ABSTRACT Flaw indications were detected in five welds of the main steam piping system.

The indications were primarily slag inclusions and subsurface, although a lack-of penetration imperfection was also observed. In assessing the significance of these weld indications in main steam piping, acceptance standards were reviewed and the ultimate capacity or residual strength under monotonically increasing.

load was calculated. The margin of safety was also determined and compared with evaluation acceptance criteria of Section XI of the ASME Boiler and Pressure Vessel Code. Based on this investigation, it was concluded that the flaw indications would be acceptable under the Summer 1978 standards of Section XI.

The results of a conservative flaw evaluation show that presence of the flaws have a negligibly small effect (< 5%) on the static strength of pipe and that leak-before-break is assured.

TABLE OF CONTENTS Section Page ABSTRACT i

NOMENCLATURE iv 1

INTRODUCTION 1-1

1.1 Background

1-1 1.2 Scope of Report 1-2 References 1-3 2

INSPECTION OF MAIN STEAM LINES 2-1 2.1 Introduction 2-1 2.2 Radiographic Inspection 2-1 2.3 Ultrasonic Examination 2-4 2.4 Radiographic Image Enhancement 2-4 References 2-7 3

FLAW ACCEPTANCE 3-1 3.1 Acceptance Standards for Planar Indications 3-1 3.2 Flaw Evaluation Acceptance Criteria 3-4 References 3-6 4

METHOD OF APPROACH 4-1 4.1 Failure Behavior of Flawed Pipes 4-1 4.2 Evaluation Strategy 4-3 4.3 ASME Section XI Flaw Evaluation Procedure.

4-5 4.4 Acceptance Criteria for Flawed Pipes 4-5 References 4-7 5

ANALYSIS INPUT INFORMATION 5-1 5.1 Introduction 5-1 5.2 Flaw Models 5-1 5.3 Material Properties 5-3

Table of Contents Tcontinued)

Section Page 5

ANALYSIS INPUT INFORMATION (continued) 5.3 Material Properties (continued) 5.3.1 Determination of K from COD Values 5-3 a

Ic 5.3.2 Determination of KIc from Charpy V-Notch Values 5-6 5.3.3 Experimental Work on E7018 Root Toughness 5-8 5.3.4 Summary of Toughness Evaluation 5-9 5.4 Stress Data 5-11 References 5-13 6

FRACTURE ANALYSIS 6-1 6.1 LEFM Analysis 6-1 6.2 EPFM Analysis 6-8 References 6-9 7

LIMIT LOAD ANALYSIS 7-1 7.1 Limit Load Model 7-1 7.2 Numerical Results 7-3 8

CONCLUSIONS 8-1 CONTROLLED. DOCUMENTS REFERENCE LIST (CD)

A-1 CONTROLLED DRAWINGS REFERENCE LIST (CDD)

A-2

-iv NOMENCLATURE Symbol Definition a

Crack depth as Critical crack depth c

a f Final or end-of-life crack depth a/t Crack penetration for surface flaw 2a/t Crack penetration for subsurface flaw a/Z Crack aspect ratio at Through-wall crack half-angle 6

Crack opening displacement (COD) s cCritical crack opening displacement E

Modulus of elasticity Distance from mid-thickness centerline to crack center 2e/t Crack eccentricity E

Applied strain Cy Yield strain y

fa Factor-of-safety based on flaw size f

Factor-of-safety based on load p

I Moment of inertia K

Stress intensity factor AK Range in stress intensity factor K Ia Crack arrest toughness

'Ratio of fracture toughness to the critical flow stress multiplied by /f c

Distance from neutral axis D 0Outside pipe diameter da/dN Crack growth rate K Ic Plane strain fracture toughness Crack length Poisson's ratio P cCritical applied load p

Applied load Shift in neutral bending axis

-V Nomenclature (continued)

Symbol Definition Applied Stress Gb Applied bending stress a

Critical stress a kLimit or flow stress om Applied membrane stress Yield strength outs Ultimate tensile strength t

Pipe wall thickness

1.0 INTRODUCTION

1.1 Background

During inservice inspection of circumferential weld seams on the 24-inch main steam piping at San Onofre Nuclear Generating Station (SONGS) Unit 1, i dications were detected in five welds by radiographic techniques (RT).

The base pipe is ASTM A-106 Grade B steel and the piping system was fabricated with E7018 weld rods.

The indications were identified primarily as slag inclusions located subsurface and in the weld metal.

In one weld, a surface-connected,planar,lack-of-penetration defect was also observed at the pipe inside surface. The detection of these defects by RT was confirmed by ultrasonic examination (UT), and flaw size information was recorded. In all cases, these indications are believed to be pre-existing defects from fabrication and did not grow during the fourteen years of plant operation.

The inservice inspection requirements for the plant are according to the -1974 Edition, Summer 1975 Addenda to Section XI of the ASME Boiler and Pressure Vessel Code (1-1), and repair is to be performed by the procedures in the Summer 1978 Addenda to the 1977 Edition of the Code. Although standards for allowable planar indications in ferritic materials are provided in the Summer 1975 Code, these standards are only for safe-end attachment welds and no acceptance standards are given for piping.

A tech Engineering Services was contacted by Bechtel Power Corporation and Southern California Edison (SCE) engineering staff to review these results and to evaluate the defects according to Code rules, where applicable.

A judgement can then be made to either defer repair to a future time or possibly show that the presence of these defects will not affect operational safety for the remaining service life of the plant.

W 1-2 1.2 Scope of Report This report is an assessment of the significance of these weld indications to the safe operation of the plant. This assessment is based on (1) a review of the Code acceptance standards for piping subsequent to the Summer 1975 Adden da, and (2) a flaw evaluation analysis to estimate conservatively the critical flaw size and critical loads to cause failure consistent with the procedure in Appendix A of Section XI. A fatigue analysis to predict subcritical flaw growth under the assumption that.the defects are crack-like has not been performed at this time. The fact that these defects were introduced during fabrication and have not been observed to grow during service suggests fatigue is not a problem and an analysis to predict final flaw size due to applied cyclic loads will be performed later, if required.

• 1-3 REFERENCES 1-1. ASME Boiler and Pressure Vessel Code,Section XI, "Rules for Inservice Inspection of Nuclear Power Plant Components," 1974 Edition, Summer 1975 Addenda.

mm

(

2-1 2.0 INSPECTION OF MAIN STEAM LINES 2.1 Introduction During inservice inspection of circumferential weld seams on the SONGS Unit 1 main steam lines, five welds were found to contain indications of sufficient sIze to warrant analysis. The screening inspection was performed with radio graphic inspection techniques, and those defects that were rejected were subsequently examined ultrasonically. In addition, the radiographs were examined using digital enhancement techniques.

2.2 Radiographic Inspection A radiographic examination of the main steam line was performed as specified in Section V of the ASME Boiler and Pressure Vessel Code (2-1). The preservice criteria for acceptance of flaw indications is Paragraph NC5320 of Section III of the ASME..Code (2-2).

Based on these preservice inspection requirements, five welds were determined to have unaccentable flaw indications. These five welds are illustrated in Figure 2-1 from main steam line drawings (CDD-1).

The flaws were generally determined to be slag inclusions; however, one lack-of-penetration defect was identified. Those flaws that were determined to be unacceptable are summarized in Table 2-1. The locations tabulated are measured from top-dead-center clock-wise around the.circumference, viewed in the direction of flow. It should also be noted that the flaw lengths tabulated are for groups of flaws, rather than individual flaws, and the welds were generally rejected due to the proximity between flaws rather than the size of individual flaws.

6o-

t 1, U,

II I I

IZ' HORIZ. P/PE PESTR.

/4 o vE pr PI PE StJ PPT

'a L3OT-e'.I,'rso13T 64 1941 Figure 2-1. Main Steam Lines -

Steam Generators to Penetrations (from Reference CDO-1).

Welds containing defects are marked with arrows.

2-3 TABLE 2-1

SUMMARY

OF RADIOGRAPHIC EXAMINATION Weld Flaw Type Flaw Location 1-04-3 Slag Inclusions 442"-45-"

Slag Inclusions 50"-56" 1-24-5 Slag Inclusions 47"-49" 2-24-2 Slag Inclusions 14"-19".

Slag Inclusions 22"-24" Lack of Penetration 64"-66" Porosity 72"-74" 2-24-3 Slag Inclusions 14"-18" Slag Inclusions 28"-30" Slag Inclusions 38-43" Slag Inclusions 46"-48" Slag Inclusions 50-52" Slag Incl usions 73"-76" 6-24-2 Slag Line 22"-30" Measured from top-dead-center clockwise around circumference of pipe viewed in the direction of flow.

Ai

2-4 2.3 Ultrasonic Examination Those defects that had been rejected during the radiographic examination were also examined ultrasonically to size the flaws. The ultrasonic inspection was calibrated using SCE's calibration block No. 40. The.

calibration block is a 4" x 7" x 0.980" block of SA516 steel containing a 3/32" side-drilled hole (ASME Section V) and a 10% sawtooth notch (ASME Section XI) as reference reflectors. The primary reference was the 3/32" side-drilled hole.

The ultrasonic inspection results (CD-14) are summarized in Table 2-2.

The depth of the flaw listed in Table 2-2 is defined as the size of the flaw in the through-thickness direction, whereas the length and width of the flaw are defined as the flaw dimensions in a plane parallel to the pipe free surface. (Note that flaw depth in the UT Report is distance below OD surface and is not to be confused with the definition in Table 2-2).

All of the slag inclusion flaws lie on this plane, and are therefore actually laminar defects.

2,.4 Radiographic Image Enhancement In order to characterize better the weld defects, the radiographs were enhanced with digital imaging and computer analysis. In order to do this, the radiographs are illuminated from behind and viewed with a low light TV camera. The image is then digitized, where each discrete bit of information is characterized by one of 2000 levels of gray. This is in contrast to the.

32-64 gray levels which can be resolved by the eye. Small density differ ences which cannot be seen by the eye can be viewed by expanding a small range of gray to the full white-to-black range. This is easily done by a computer because the data are in binary form.

All welds containing rejectable defects were analyzed by this method (CD-13).

Figure 2-2 shows both a non-enhanced and an enhanced picture of the slag inclusion cluster at 47"-49" in Weld 1-5. From this evaluation it is clear from these pictures that the flaws are confirmed to.,be clusters of small inclusions, all of which would be acceptable on an individual basis.

[~~~~~

V V

TABLE 2-2

SUMMARY

OF ULTRASONIC INSPECTION DATA*

Depth Below Weld Type of Flaw DAC Len th Depth**

Width Surface in in (in)

TF7n 1-24-3 Slag Inclusion Cluster 50%

1.0 1.0 0.200 Slag Inclusion Cluster 50%

1.0 1.0 0.200 1-24-5 Slag Inclusion Cluster 30%-50%

1.0 1.0 0.200 2-24-2 Slag Line 20%-30%

5.0 0.100 0.125 Slag Line 20%-30%

2.0 0.100 0.125 Lack-of-Penetration Masked 2.0 0.060+0.010 ID surface 2-24-3 Slag Line 20%-30%

38.0

-0.100 0.125 Slag Line 20%-30%

3.0 0.100 0.125 6-24-2 Slag Line 20%-30%

8.0 0.100

.0.125

• All flaws evaluated at 28dB primary reference at 1/2 Vee path. Attenuation is 4dB maximum.

• • For analysis purposes, the depth of slag inclusions is assumed to be 1/16 inches.

2-6 a) non-enhanced SLAG INCLUSION CLUSTERS b) enhanced Figure 2-2.

Example of Digital X-Ray Image Enhancement.

2-7 REFERENCES 2-1 ASME Boiler and Pressure Vessel Code,Section V, "Nondestructive Examination".

2-2 ASME Boiler and Pressure Vessel Code,Section III, Division, Subsection NC, "Class 2 Components".

3-1.

3.0

-FLAW ACCEPTANCE 3.1 Acceptance Standards for Planar Indications The acceptance standards for flaw indications detected during inservice tnspection are covered in IWB-3000 in Section XI for Class 1 components.

In the 1974 Edition of the Code and in the present state of development, the Class 1 requirements are applicable to Class 2 and 3 systems.

As introduced earlier, there are no explicit acceptance standards for.

piping in the 1974 Edition, Summer 1975 addenda to Section XI.

However, guidelines for the acceptance of planar indications in nozzle-to-safe-end attachment welds are discussed and the Code refers to the standards for pressure vessel welds (Table IWB-3351-1) for ferritic steels with thicknesses of 4 inches or greater.

This thickness requirement makes the application of these standards to piping very conservative, and a comparison of the detected flaws with these standards shows a non-acceptance situation for both the buried and surface flaw indications as shown in Figure 3-1.

Revisions have been made to the inspection standards, and flaw acceptance limits for piping have been introduced.

The acceptance standards in the 1977 Edition, Summer 1978 Addenda from Table IWB-3514-2 are also plotted in Figure 3-1.

All flaw indications are acceptable under the Summer 1978 requirements.

A summary comparison of actual flaw sizes with allowable sizes is given in Table 3-1.

The allowable flaw penetration.for the buried flaws exceeds the conservatively assumed values for the depth of slag inclusions by almost a factor of four.

The lack-of-penetration defect also satisfies the standards for surface indications by a factor of one and one-half.

This comparison indicates that there exist significant margins in the flaw measurements which can cover uncertainties in the inspection method, on flaw sizing.

tH AIR L I- -L 4-4 T1. 41 4-

- I-,

F

-ja T-F.

+

TTIT 4-1:.

T.

I M

71 HH4;4- *:

J I-rr t

T.-

T 4

M4 J,

Fr

i I iT.

1 1-TH C"CW-7. 4 T i4 t4--Hr-11 T.11; 11:

-.4

-17 Tw T

1. 44-7-1 f-J_4i t4 -#t-A-h.

T4 TIR t

H-;

r" T V T

11 LA I r

i 1111 (Y) 4 1

I ITTT I t

60 4

lit..

-T i

+

T! T

- T

-1 1:

A ;

14- _1: -1

+ -

r

1411, 4-f-4A+/-

t 4

T I

T D:

+

4-1 it 4##

14T

-1 T

TTT 4

T]

T 7-77' W

I -T. 1-14 7

w 7

iT IT iij J

T TIT

'4 TH BIT UF T TH p m.

f PIT' 7

],

1

J TI-1 11 1 '17,

4 L7rT

.4 zlr 41-q-.-

A'4

-P T.

AH T-'

T T

-77 T;

!-i IT;;

1:

1: iq l Trl y

6iw

1

-T-T f f -H T,

_j T

4-p-J-1 14 X1-1J, T

TZ III#

la-F woo 77

-1 7

T 4-;

-T.

t:

7 7:

j" A

iT -7 IT 77i Im fli 77 b.0 ttt Z-i T -7

-.' -tt T

. 1 !

3-3 TABLE 3-1 COMPARISON OF FLAW SIZE DATA WITH

.SUMMER 1978 ACCEPTANCE STANDARDS Ratio of Measured Measured*

Allowable Allowable to Weld No.

Type a/i.

a/t,(%)

a/t,(%)

Actual 1-3 Subsurface 0.031 3.3 12.8 3.9 1-5 Subsurface 0.031 3.3 12.8 3.9 2-2 Subsurface 0.006 3.3 12.6 3.8 Subsurface 0.016 3.3 12.7 3.9 Surface 0.035 7.2 10.6 1.5 2-3 Subsurface 0.001 3.3 12.6 3.8 Subsurface 0.010 3.3 12.7 3.9 6-2 Subsurface 0.004 3.3 12.6 3.8 Subsurface flaw size (2a) for slag is assumed to be 1/16 inches.

A requirement for Class 1 ferritic steel is that it meets the Charpy impact requirements of NB-2300 for piping. That requirement for A-106 piping over 3/4-inch thick but less than 1.5 inches is 25 mils lateral expansion 25 ft-lbs).

Although no guidance is given by the Code on how to apply Class 1 standards to Class 2 piping when the material was not purchased to a minimum Charpy energy level, lower bound Charpy energy data for base and weld metal will meet this requirement. This information is presented later in Section 4.

3.2 Flaw Evaluation Acceptance Criteria Although the flaws are acceptable under the later version of the Code used by SONGS 1 for repair, the defects were still evaluated under the guidelines of Appendix A of Section XI.

Flaw indications exceeding the limits of IWB-3500 are acceptable if a flaw evaluation shows that ample safety factors exist either on flaw size or applied load for normal operation and for emergency and faulted conditions.

The analysis procedures for pressure vessels have been established to determine the followina flaw size oarameters:

a The maximum size to which the observed flaw can grow during the remaining service lifetime of the component a

The minimum critical size of the observed flaw under normal conditions ai-The minimum critical size for initiation of non-arresting growth (fast fracture) of the observed flaw under emergency and faulted conditions.

The flaw acceptance criteria of Paragraph IWB-3600 require that repair may be avoided only if the following margins exist:

3-5 af < ac/10, (3-1) a

< a /2.

3-2)

The above criteria (IWB-3611) require that the computed final flaw size would have to be less than one-tenth of computed critical size for normal conditions and one-half the critical flaw size for emer gency and faulted conditions. The former criteria, Eq. 3-1, can be technically unsatisfying since it imposes a severe geometric limitation for thinner vessels where ten times af can be greater than the wall thickness. For this reason, alternative criteria (IWB-3612) have been established based on stress intensity (or load) rather than crack size (3-1):

K

< KIa /V0 (3-3)

K

< K //,

3-4)

I.

Ic where KI is the applied stress intensity factor, and KIa and KIc are material properties which quantify resistance to fracture. These parameters and analysis procedures are discussed in Section 4.

3-6 REFERENCES 3-1 ASME letter dated June 11, 1974 (ASME File #BC-74-188).

mE

4-10 4.0 METHOD OF APPROACH 4.1 Failure Behavior of Flawed Pipes The failure behavior of piping materials containing defects and under Tonotonic loading can be classified into three regimes in which a specific type of failure mode is appropriate. The disciplines required to assess these regimes are:

(1) Linear Elastic Fracture Mechanics (LEFM)

The structure fails in a brittle manner and, on a macroscale, the load to failure occurs within nominally elastic loading.

(2) Elastic-Plastic Fracture Mechanics (EPFM)

The structure fails in a ductile manner, and significant stable crack extension by tearing may precede ultimate failure.

.(3)

Fully Plastic Instability or Limit Load The failure event is characterized by large deflections and plastic strains associated with ultimate strength collapse.

A diagram showing the relationship between the critical or.failure stress hnd flaw size for the three failure modes is shown in Figure 4-1. The shape and position of the failure locus will depend on the fracture toughness (KIc) and strength properties (a and a uts) of the material, as well as the structural geometry (t).and type of loading. LEFM is used most appropriately to describe the behavior of low toughness/high strength materials in which the plastic zone is small relative to the structural geometry and little ductility precedes fracture. With this method, no account, is taken of increased material resistance to brittle fracture when significant plasticity occurs.

Under LEFM conditions, the most useful parameter for characterizing the behavior of cracks is the stress intensity factor, K, which characterizes the singular stresses near the crack tip.

ta lIil BBaBLAIBAa Limit Load Elastic-Plastic (EPFM)

Fracture (LEFM) mLimit Load Limitin FaFluecMode

4-3 In contrast, plastic instability, when it occurs without prior crack extension, is.dominated by the flow properties of the material. In these circumstances, the.failure condition is independent of fracture toughness and crack tip characteristics, and a limit load analysis is used to define the failure condition. Elastic-Plastic.Fracture Mechanics (EPFM) analysis can be used to predict failure behavior in the transitional regime between LEFM and limit load, and under EPFM conditions, the crack tip singularity, the material toughness., and net section strength are all important parameters for failure assessment. EPFM principles which incorporate a crack opening displacement.

(COD) concept have been applied to predict failure loads under elastic-plastic conditions.

4.2 Evaluation Strategy The strategy for the evaluation of weld defects is outlined in Figure 4-2.

Material properties information and the applied stresses are used in the analysis to compute the critical flaw size, ac, for the pipe weld. The material properties used in the analysis are the fracturetoughness of the material quantified by the critical stress intensity factor, KIc' or the critical crack opening displacement, 6c* Also important will be the yield (y ) and ultimate tensile (;uts strengths at operating temperature.

The applied stress is assumed to be a combination of uniform tension (a )

m and bending (ab) stress fields arising from pressure, dead weight, thermal expansion (including cold pull), and seismic loadings.

The computations to determine ac are based on LEFM, EPFM and limit load methods, and the smallest value of ac will be used for comparison with acceptable flaw evaluation limit. With a knowledge of actual flaw size,the factors of safety based upon flaw size (fa) and applied load (f ) are determined. Based on the a

p safety factors present and criteria for flaw acceptance consistent with Section XI, Appendix A, the significance of the defects will be determined.

4-4

-Materials Stress Inspection Data Analysis Data K

6 y

uts m b Critical Flaw Size C

Factor of a, 1,e Safety fa' p Flaw Acceptance Criteria Figure 4-2. Flaw Evaluation Strategy for Fracture.

4-5 4.3 ASME Section XI Flaw Evaluation Procedure The flaw evaluation procedure for determining the acceptability of flaws W ithat exceed the allowable flaw indication standards is contained in Appendix A of Section XI (4-1). Appendix A is considered non-mandatory in that the evalua tion of detected flaws need not be performed if repair is made. Although the flaw evaluation procedures in Section XI are based upon the principles of linear elastic fracture mechanics and are intended for thick-wall structures such as pressure vessels, the stress intensity factor solutions and the fatigue analysis procedures for calculating the final flaw size, af) can be used in this assessment of piping. In the fatigue analysis requirements, the flaw is grown in geometrically similar sizes until all expected service cycles have expired.

The complete evaluation approach is described in Article A-5000 Section XI, and a computer program called FACET (4-2) has been developed which uses the Section XI procedure. FACET is used in this study to compute the critical flaw size according to LEFM for the defects. The background and development of these procedures in Section.XI have been recently documented (4-3).

4.4 Acceptance Criteria for Flawed Pipes Although a flaw acceptance criterion for defects in piping which exceed the allowable standards does not exist, a set of conditions can be established which reflect the intent of Section XI, Appendix A, specifically the flaw acceptance criteria IWB-3600 (4-1). The Code requires a check on.either flaw size, or applied K (or load) as presented in Section 3.2 to establish whether a flaw can remain in a component. The two conditions can be written as:

a Sf a' (4-1) 4f K1

_Ic o c f

(4-2)

K I Ic K

r--

a

4-6 where a, K and a are the critical values for flaw size, fracture toughness, c

Ic c

applied stress and af is the final flaw size assumed to be equal to the detected flaw size (i.e., neglecting fatigue crack growth), a is the applied load, and fa and f are the calculated factors of safety on flaw size and applied load. It should be noted that f and f in general are.

a p

not the same and may, in fact, have different values and still provide the same assurance against failure. In the Section XI procedure, the acceptance criteria are focused on assuring that:

f

> 10,

(4-3) a f

> /TU (4-4) p for normal operation, and for emergency and faulted conditions fa > 2, (4-5) f >

4-6 p.

4-7S REFERENCES 4-1 ASME Boiler and Pressure Vessel Code,Section XI, "Rules for Inservice Inspection of Nuclear Power Plant Components," 1977 Edition.

4-2 Cipolla, R. C., "FACET:

Computational Method to Perform the Flaw Evaluation Procedure as Specified in the ASME Code,Section XI, Appen dix A," EPRI NP-1131, September 1979.

4-3 Marston, T. U. (ed.), "Flaw Evaluation Procedures -- Background and Application of ASME Section XI Appendix A," ASME Task Group on Flaw Evaluation, EPRI NP-719-SR (August 1978).

5-1 5.0 ANALYSIS INPUT INFORMA ION 5.1 Introduction To perform an accurate flaw evaluation, the interaction of three parameters must be defined. These are:

the geometry of the flaw and structure, the appropriate material properties which describe the resistance of the structure to failure, and the applied stresses. Given the correct physical model, once any two of these input parameters are known, the third can be quantified. The parameters necessary to perform the fracture analyses are presented next.

5.2 Flaw Models The radiographic and ultrasonic examinations have shown that there are basically three types of flaws of concern:

slag inclusion clusters, slag lines, and lack-of-penetration. Thus, for the fracture analysis, three flaw models will be assumed.

The lack-of-penetration defect is a planar defect oriented normal to the longitudinal axis of the pipe.

It is also connected to the inside surface of the pipe. Thus, it will be modeled as a semi-elliptical surface flaw of length, Z, and depth, a, as shown in Figure 5-1.

The slag inclusion defects are laminar defects (they lie in a plane parallel to the pipe free surface), which are buried 0.125 inches and 0.200 inches below the surface. Because of their orientation, the depth, a, is more important than the width, w. However, the depths of the flaws were not determined by the UT examination. Thus, for analysis purposes, the depth of a layer of slag has been assumed to be 1/16 inch.

This value wa§ shown to b'e an upper bound on slag inclusio6 depth for welds in other pipe liine applications (5-1).

5-2 OD surface a

ID surface Figure 5.1 Schematic Diagram of Flaw Models

5-3 The slag inclusion clusters are modeled as buried flaws (0.200 inches below the surface) of length, k, and depth, 2a, as shown in Figure 5-1.

The slag lines are also modeled as buried flaws (0.125 inches below the surface) of length, Z, and depth, 2a.

Table 5-1 summarizes the lengths, depths, and eccentricities of all flaw models.

It should be noted that the dimensions tabulated in Table 5-1 for slag inclusions are the dimensions of the envelope that contains the group of individual slag inclusions rather than individual flaw dimensions.

In addition, the flaws are assumed to be sharp cracks.

Both of these assumptions ensure that the analysis performed will be conservative.

5.3 Material Properties The flaws of interest are contained in welds made by the shielded metal arc process with E7018 electrodes. The base material of the pipe is ASTM A-106 Grade B carbon steel. In the evaluation of properties presented in the following sections, properties of the weldment are assumed to be the lesser of the weld metal or base metal properties. This will result in a conservative value for toughness.

Based on minimum specified tensile strengths, the base metal strength proper ties are the more conservative.. The mechanical properties of A-106 steel,.as specified.in the ASME Code (5-2), are used in this analysis. Although the pip ing system was designed to ANSI B31.1 Standard for Power Piping (5-3), the use of ASME Code values is in accordance with the B31.1 design standard which requires that materials for boiler external piping be specified in accordance with the applicable SA specification of the ASME Code. The mechanical properties of SA-106 steel are summarized in Table 5-2.

5.3.1 Determination of KIc from COD Values A literature search was performed (5-4) to determine the fracture toughness values appropriate to this application. This search provided both critical crack opening displacement (COD) and Charpy V-Notch values as determined in

TABLE 5-1

SUMMARY

OF FLAW MODELS Depth Penetra-Aspect Eccen Weld Length Depth*

Below tion* a/t Ratio tricity Number Type of Flaw k (in) a or 2a (in Surface (in or 2a/t a/Z 2e/t 1-3 Slag Cluster 1.0 1/16 0.200 0.065 0.031 0.587 1-5 Slag Cluster 1.0 1/16 0.200 0.065 0.031 0.587 2-2 Slag Line 5.0 1/16 0.125 0.065 0.006 0.742 Slag Line 2.0 1/16 0.125 0.065 0.016 0.742 Lack-of-Penetration 2.0 0.070 0.0 0.072 0.035 2-3 Slag Line 38.0 1/16 0.125 0.065 0.001 0.742 Slag Line 3.0 1/16 0.125 0.065 0.010 0.742 6-2 Slag Line 8.0 1/16 0.125 0.065 0.004 0.742 use a for lack-of-penetration flaw, 2a for all other flaws.

5-5 TABLE 5-2 MATERIAL PROPERTIES OF SA-106 GRADE B STEEL PIPE 540oF Room Temperature Yield Strength (ksi) 27.3 35.0 Tensile Strength (ksi) 60.0 60.0 Modulus of Elasticity (ksi) 26.1 X 103 27.9 X 103 Possion's Ratio, v 0.3 0.3

5-6 experimental work. The results of this survey are shown in Figure 5.2. The most conservative values found have been utilized to assure conservative K evaluation. A toughness level based on elastic IC plastic.fracture mechanics is considered to be most appropriate to this material, stress level and geometry. The plane strain fracture toughness can be found from elastic-plastic theory using the relationship (5-5):

K 6

Ic c

(5-1)

'y )

y where a here is taken as the elevated temperature yield stress of A-106 Grade B, c

is the yield strain, and 6 is the critical COD value. The literature search y

c indicated a range of 6 values for welds like E7018 to be 0.023 inches to 0.052 inches at 32oF (5-6).

For higher temperatures such as the normal operating temperature of this piping system the critical COD values (and thus, KIc values) will be substantially higher. The use here of 6c = 0.023-resultsin conservative, lower bound KIc values. Note as well that the welds reported in the literature were not post-weld heat-treated; since the SONGS welds were post-weld heat-treated, they will be characterized by improved toughness, as discussed in the next section.

The use of 6c = 0.023 inches and equation (5-1) provide a value for K

= 97 ksil'ln.

This value is used in the calculations that follow in Section 6 to analyse critical flaw sizes to cause fracture. As confirmed in the next two sections this results in a conservative measure of the structure's capacity to resist fracture initiation.

5.3.2 Determination of KIc from Charpy V-Notch Values As a ch.eck on the value for K obtained above, an LEFM-based fracture IC toughness was calculated based on Charpy V-notch values and a commonly applied empirical correlation. The.appropriate data were taken from the literature search (5-6,5-7) and are shown in Figure 5-2.

The lowest curve (5-8) is for base metal specimens oriented circumferentially, where the notch runs parallel to the longitudinal axis of the pipe. As the defects are

im~

I t

T-T

'AFF1T'-

mI[L It4!

1

-4+1 TT i}Fttfh. -.,-

4TH1~

i [

i#

-i~ 4 ir i

t-F I-it.-:

i-

-114 -

4 F I n

1 I

T, T

I_ :jil1f t

1 T

p rillilK

+i i F :I:. :l :T Ft I

T 0

ii-H H i J

ITl I-T-l :- -I 14 I

F I rr EI'-F7 C')

.F....FF II' 444 1 2

~

F' I.

I1 4~ TH F~

T H -HI ~

I 6#

F2r1 rjB g

TI, tF

/

=T_

F

.. 4 L.4-1I4.F II i

F, i

Th lh IiliH F:

I ~~*'

THT ti_

7-iI t ':

I~4

!III 41 J9-'

I 4

A J-1 F1i{

Ti.

4 It 7:1-- -----

If II'Ft A

5-8 oriented circumferentially, and are located in the weld metal, this was not considered an appropriate measure of toughness. The next lowest set of values are those for the root pass of as-welded weld metal.

This curve was used as a valid lower bound on Charpy values for the present application. Note that other weld regions, away from the root, (where all defects except the lack-of penetration defect are located) will have higher inherent toughness.

Note as well that data from the same sources (5-7) indicated an improvement on Charpy properties for materials that are post weld heat treated.

Charpy values can be converted into KIc values using the Rolfe-Novak-Barsom correlation (5-5) for upper shelf Charpy values:

2 K6 2 IC 5

(CVN -

)

(5-2)

CY G

20 y

y where the yield stress a, is evaluated at the same temperature as the Charpy test. This equation has been shown to be appropriate for high strength base materials. No such correlation exists for as-deposited weld metals and the method is used here to establish a guide to the lower bound toughness.

Using the lower bound root toughness curve, the upper shelf fracture toughness of the material is estimated to be 97 ksi/in. This corresponds well with the EPFM-based fracture toughness calculated earlier.

5.3.3 Experimental Work on E7018 Root Toughness Experimental work to determine actual dynamic fracture toughness values on 1" thick carbon steel plates with E7018 welds and no post weld heat treatment has been performed elsewhere (5-4).

The results of that particular investigation show substantially higher toughness values and definite upper shelf behavior for test temperatures greater than 500F. These data confirm the conservative nature of the preceding sections.

5-9 5.3.4 Summary of Toughness Evaluation A summary of appropriate toughness information is shown in Table 5-3. It should be emphasized that the following conservative assumptions have been made in this calculation of K :

1) The K values shown are calculated from COD and CVN values taken at 32oF and 158 0 F respectively.

For normal operation at 540 0F, higher toughness will result.

For lower temperature operations, the applied stresses will be much lower than discussed in the next section.

2) The CVN values used were found for the root pass in welds equivalent to E7018, without post weld heat treatment (PWHT).

As shown above, the use of PWHT will result in higher toughness levels.

3.) All flaws except the lack-of-penetration defect are in weld locations outside the root-areas which have inherently higher toughness.

4) The lowest material properties (for example base metal vs. weld metal, high temperature vs. low temperature) are used which gives the lowest equivalent toughness from the empirical relationship used.

5) Actual measured toughness values for the root region in E7018 welds on a 1" base plate of similar material give substantially higher values than the toughness values determined from empirical relationships for COD or CVN data, even though these specimens were not post weld heat treated.

5-10 TABLE 5-3

SUMMARY

OF TOUGHNESS VALUES Test Equivalent Post Weld CVN Value Temperature KIc Region Heat Treated?

(ft.lb.)

(OF) ks i vir Weld root no 58 158 0 F 97 (lower bound)

Weld root yes 95 104oF 113.05 (lower bound)

Weld root no

> 160 ft.lb.

> 500 F 300 (KId) entire weld no

.023* through 32oF 97

.052 This, is a COD value in inches.

5-11 5.4 Stress Data The stresses used in the analyses in Sections 6 and 7 are based on stress values obtained from Bechtel (CD-12).

The load cases considered include pressure, thermal expansion including cold pull, seismic, dead weight, and a turbine stop valve closure. Table 5-4 summarizes the stress information.

The load cases include normal operation, upset and emergency conditions, and the total. stress listed in Table 5-4 represents a worst case situation since all forms of loading are summed together. As all welds are stress relieved, residual stresses due to welding are assumed to be negligible.

TABLE 5-4

SUMMARY

OF STRESSES Turbine Weld Pressure Dead Thermal Stop Valve Total Membrane Bending Number Stress Weight Expansion Seismic Closure Stress Stress Stress (ksi)

(ksi)

(ksi)

(ksi)

(ksi)

(am+ob)

(am)

(ob)

(ksi)

(ksi)

(ksi) 1-3 6.2 0.47 8.48 1.17 7.3 23.62 6.40 17.22 1-5 6.2 0.55 10.09 0.92 2.7 20.46 6.74 13.72 2-2 6.2 0.17 10.18 0.84 2.8 20.19 6.87 13.32 2-3 6.2 0.48 8.91 0.81 7.3 23.70 6.40 17.30 6-2 6.2 0.27 7.20 2.26 7.7 23.63 6.81 16.82

5-13 REFERENCES 5-1. National Bureau of Standards, "Consideration of Fracture Mechanics Analysis and Defect Dimension Measurement Assessment for the Trans Alaska Oil Pipeline Girth Welds," PB-260 401, October, 1976.

5-2.

"ASME Boiler and Pressure Vessel Code,"Section III, Division I Appendices.

1977 Edition, including Summer 1979 Addenda.

5-3. American National Standards Institute, B31.1, "Power Piping."

5-4.

McNaughton, W. P. and G. R. Egan, "Analysis of the Significance of Weld Gaps Between Weld Backing Bars in Perry Nuclear Station,"

Aptech Engineering Services.Report AES-79-07-06, July 1979.

5-5.

Rolfe, S. T. and J. M. Barsom, "Fracture and Fatigue Control in Structures," Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1977.

5-6. Nordell, W. J. and W. J. Hall, "Two Stage Fracturing in Welded Mild Steel Plates," Welding Research Supplement, March, 1965, pp. 124s-134s.

5-7. Robinson, J. L., "A Study of the Factors Controlling Root Run Toughness in Multi-Pass Manual Metal Arc Welds in C-Mn Steels," Abington, UK, The Welding Institute.

5-8. Kiss, E., J. D. Herald, and D. A. Hale, "Low-Cycle Fatigue of Proto type Piping," GEAP-10135, January 1970.

6-1 6.0 FRACTURE ANALYSIS 6.1 LEFM Analysis The ASME Section XI flaw evaluation procedure is based on the principles of linear elastic fracture mechanics, and the complete evaluation procedure is described in Article A-5000 of Section XI.

The stress intensity factor, K, is calculated based on the applied stresses and flaw geometry. Under LEFM assumptions, unstable fracture is initiated when the value of K reaches a critical value, KIc* The Code procedure for calculating K has been programmed into the computer code FACET (Flaw Analysis and Code Evaluation Technique) (6-1).

FACET was used to calculate the stress intensity factor, K, for a uniform stress field for three flaw geometries: a surface flaw on the pipe ID, a buried flaw 0.200 inches below the OD surface, and a buried flaw 0.125 inches below the OD surface. A uniform stress field was assumed because the through-thickness variation in bending is small in a 24-inch pipe. This assumption is conservative if the uniform stress equals the peak stress (membrane stress plus bending stress).

With a knowledge of the applied stresses as summarized in Section 5, the stress intensity factor can be determined for each flaw as a function of flaw size. These calculations are summarized in Figures 6-1 through 6-3.

Entering these curves at the flaw sizes as determined in Section 5, the applied stress.intensity factors can be determined for each flaw. These calculations are summarized in Table 6-1 for each flaw. Note that the applied stress intensity factors are significantly (an order of magnitude) smaller than the fracture toughness, KIc. This indicates that the stresses could be ten times higher before fracture would occur. From Figure 6-1, a flaw must be 0.57 inches deep before fracture of the uncracked ligament would occur under given loading. However, the entire pipe section would not fracture (i.e., the crack would be a leaking crack).

The pipe could support a through-thickness flaw approximately fifteen inches long (Figure 4-4) before fracture would occur under the given loading, based on linear elastic fracture mechanics.

r~~~~i rtvirF V

I I

F L

P X0f TO 46132

........i

+-t j

j

AJ H

IT-f

+/-f 4-4-+4 01 4:'T 4-,4

+H+.

IT I4.

2-~l

-p'

+d 44-7 jj -

14-z

I_

_ _ __ _ _ TU'

-77 J777 1

4 t

Iffl, H2-

-44r-7 IT

--Hill AN

-H K--

__7___

_7

FUEF A~ ES=CC.

46 1322 110

-1 T--

• -I TI 4-- 1 4

T 4-4 -71 t

2 1-T H

4 j 4A j-4 4'4 7_

4 4t IT_

1.

i1.4V

-1j.4 4

I

6-

.. - --I*

ITT T

1 TT TITT 4-1+

l4- -41t S

-i I

r'-

, UTi4 fl+l-4

-H'++

t IT Ad# T -

6w J

4- +

'T

-HiT, 4

+

4 jt L ETU_

H-H H H-t-1 +111 T Ej: 'i;-; t'

-P74 fl M "M T"T T -I I I T

"-r I

T

_J I.

F I

+1-7

-t

-4 r

f4 -I T _114

'T -PT T

T" f-I T

60

'77

+

T

"'P TI:

4-f;

-TH 7!l H

-T- -l;-F T

+H- -H +

j44+

+

!-H-r i

T j

A 7P

_CT 71

+

ttt!

TIL

-4, I it 1 11 IT Z

T_

-H

_Hr T

H-if 4

t Hj T

T17 4-T: I i T

T4 H7 7_7T_

t-I i j-I-

tH 00 Tri Y,

W tF I i T

Au OF dl

+

41 L

ilti_

1

'H1, 7.777 T-t-T 1-t i_ _ _,T _T ME T

T T I--

-7 T

i+H i

':7r 1

TT f

4

-A 1 ITF -t rr

-rT-rT-T1,7,1, 1#1 I-A+ +i ftm MT ir I.. !TT' W -1:44 7-4 "H

it 7,

TIT!

'-'_1 1- -TT iz"

, /

L 4 fll!

if'

T ti T,

"J.

J" JJ_

ttt t-,

101 E.,

T_

t Ab 4 7T T TT__

1 4 i.

-4 4-H.

141

+

T u

_4

-n

-FT i-i_:Ti D -0 Z

iTTJEFE T17 H

'qT1T:

'T it 4

+

11-i H'j rT1 f 1".

-M T'

11 CIL

[T ji, ij 71 T F jj I "Ti 71 T 7.1 It i t 4 i-1 ij i-il 41 T-1, id J

T.

+

4-i i l T;T J'.

4

+1-1-1! -1 IP1_

11f., F 4 1 f-1 TPT

.1-1-P.

f

_1 IJ 4-HT 1:4 I-P r I 71 i:i j-L H.

1+ 1_4 441-41 JA

hi i 1J j.i 1'1-1-1 Ht ill-i 774

.71-

=

7.......

7 17-77-,

144

_.Jj

'T 4:

Pir

_T 4-f' I IT JH I--.

4 4

7 M

T i i 1 7_1-=

if 4

-q "1 +1 t 2

I?~

I-V i-v r

r-F" U-I

~1

~

F

~

F 1

TABLE 6-1

SUMMARY

OF CALCULATED STRESS INTENSITY FACTORS Total Weld Type Eccentricity Length Depth Stress K

K /K ac ac/a 2e/t (in)

(in)

(ksi)

(ksi 7-)

C (in) 1-3 Subsurface 0.587 1.0 0.0625*

23.62 8.0 12.13 0.56 8.96 1-5 Subsurface 0.587 1.0 0.0625*

20.46 6.35 15.28 0.59 9.44 2-2 Subsurface 0.742 5.0 0.0625*

20.19

.6.8 14.26 0.59 9.44 Subsurface 0.742 2.0 0.0625*

20.19 6.8 14.26 0.59 9.44 Surface 2.0 0.070 20.19 11.8 8.22 0.59 8.43 238 2-3 Subsurface 0.742 38.0 0.0625*

23.70 8.1

.11.98 0.56 8.96 Subsurface 0.742 3.0 0.0625*

23.70 8.1 11.98 0.56 8.96 6-2 Subsurface 0.742.

8.0 0.0625*

23.63 8.1 11.98 0.56 8.96

• This value has been assumed Acceptance Criterion:

KIc/K > V -

boo 7

V T_

T

_r

-1 T

T

, Ct 4

4 TF1 444 1-H 4-t I L Ti-I t IT Ra 4

TT t-I 4 l T

4 -

4- -1; hT TTr

+ -- ---

60 I

+

7 I+

T J

-tv -

.. w 4 4

+

+

+ 4 60

jtl,

+

4_1 T T C\\j T

T TH tTI1 T-I-.-, T I#

IT II-T -1

_T L

_jj:

I I _.

-I t F fu: I

-7 :-7 14ir 4:4+1 4

r J

F -1 4

r T 4- -H I

I

. T H

!J t4-14 a I:

lj 74 A

4T

-H F

-I

,r.

-7

+

TEF 4

_H_7H 4+-

47 44

+H+

44

'T 7 M

7M

__r 7.

_T _1_r

+

FT q

T T-77

-T-T-T T1.

r

+ 1 TH.:

1_.- Ir

+4 4-7 -.4 IT:

r -4 J..

ri r-t- -I -,T I-T T

J-H -Hr.

j, H

-44 v

T

+1 7.,7 i 7-i

+1 T1 tl H

H i-r :4H

_1_TT T-P I-,

TT rl I

H TT 7 _T-1_7 T

T JjJ4 J

P7

-t t 11

- I I t I I flit 7

ii 4 T4 T_

4

+

j-1 ir, 4 _i' j+4' A i

+

T4: T,,

T 9,:, r I I -

H 4

1-r! ;1 2 1 I :-:

rl

• i

-1 1 - 1_1 I- -i I.

-," +Nrin Wd.- T:

rr I-

-E, H

_L LI tTl

-T T-1 19

+

7 H+f ki T1 + I H

+

T i-7-TH tltt!

+;

+i 4",

MI 7

4-,

T-1 1

7 Ti

+

LM TZ T-1 w

T:M 1 T

-I H-f

_P TT A

4-jj-i -ILJ-

'T "1.11 ji J

+jj+

.I.j til N-A I

T t

• rr

6-7 6.2 EPFM Analysis An elastic-plastic fracture mechanics analysis based on crack-opening displacement concepts was performed in accordance with the British Draft Standard Rules for the Derivation of Acceptance Levels for Defects in Fusion Welded Joints (6-2). The basis for the Draft Standard is the semi empirical COD design curve, which relates the non-dimensional COD, 4, and the ratio of applied strain to yield strain, c/c.

.The curve is defined as y

follows:

,2 c

for - < 0.5 (6-1) y y

0.25 for 2> - > 0.5 (6-2) y y

where 6

(6-3) y o = applied

COD, and a

allowable through-thickness crack half-length.

The Draft Standard contains rules for converting surface and buried defects into equivalent through-thickness flaws.

The results of this analysis for a surface flaw are shown in Figure 6-5.

For a 0.06 inch deep flaw, such as the lack-of-penetration defect in. eld 2-2, the allowable applied strain is extrapolated to be twelve times the yield strain, which is well beyond the fracture regime as predicted by this standard.

The flaw must be 0.38 inches deep before the given applied strain exceeds the allowable strain. The analysis of the buried defects shaws similar results.

A through-thickness flaw may be approximately 3.4 inches long before the given applied strain exceeds the allowable strain.

It should be noted that the Draft Standard is intended for use in design, and thus the calculated flaw sizes are allowable flaw sizes rather than critical flaw sizes.

Thus, the critical flaw sizes will actually be larger than those quoted.

[~~ ~ ~

r- ~r-r F

r IiI F

STRAIN vs CRITICAL FLAW SIZE (SURFACE FLAW) 80= 0.020 in, t=0.968 in, a/k =0.0 4

3 w

0' 0

0.2 0.4 0.0 0.8 FLAW DEPTH INCHES Figure 6-5.

Allowable Strain Levels for Surface Flaw,.Based on Elastic-Plastic Fracture Mechanics.

6-9 REFERENCES 6-1 Cipolla, R. C., "FACET:

Computational Method to Perform the Flaw Evaluation Procedure as Specified in the ASME Code,Section XI, Appendix A", EPRI NP-1131, September, 1979.

6-2 British Standards Institute, "Draft Standard Rules for the Derivation of Acceptance Levels for Defects in Fusion Welded Joints", 75/77081 DC, WEE/37, February, 1976.

L

7-1 7.0 LIMIT LOAD ANALYSIS 7.1 Limit Load Model The load required to form a plastic hinge in a thin-walled pipe with a part-through constant depth crack was determined analytically for the geometry shown in Figure 7-1. The pipe material is assumed to behave as a rigid perfectly-plastic material.

To accomodate for strain hardening, the limit stress (a ) for the material is assumed as the average flow stress:

= (ay + outs)/2.

(7-1)

For the values tabulated for a and outs in Table 5-2, a is 43.7 ksi at 540 F.

The shift in the neutral bending axis,caused by both the presence of the crack and combined tension-bendingis calculated from a relationship which satisfies force equilibrium in the longitudinal direction given below as (7-2) 2 2

where am is the applied axial stress for the uncracked section and at is the crack half angle. The requirement of moment equilibrium is satisfied through integration of the stress distribution across the section and equating with the applied bending moment to give b 2 (

=

2 sin sin a (7-3)

The simultaneous solution of Eq. 7-2 with 7-3 defines the failure locus for plastic collapse of the section.

Since in the derivation of these expressions an assumption was made which idealizes the pipe as a thin-wall cylinder, the location of the flaw relative to the

7-2 NOMINAL STRESS IN THE UNCRACKED SECTION OF PIPE NEUTRAL AXIS G

FLOW STRESS, m

STRESS DISTRIBUTION IN THE CRACKED SECTION AT THE POINT OF COLLAPSE Figure 7-1.

Limit Load Model and Geometry.

wall thickness (i.e., surface versus subsurface) is not a variable in the solution and the above expressions can be applied to subsurface flaws when 2a/t is substituted for a/t in Eqs. 7-2 and 7-3.

7.2 Numerical Results In all limit load calculations, both primary and secondary stresses are combined to provide for a conservative analysis. To investigate one extreme the limit load results for a part-through or.buried crack which is assumed completely around the circumference are shown in Figure 7-2. For the applied stresses for each weld location, the critical crack depth is between approximately 65 to 68%

of the wall thickness.

For the case of a crack completely through the thickness of the pipe, the critical crack angle for failure is shown in Figure 7-3.

The highest stressed location gives a critical crack angle (2a ) of c

0 approximately 117 or a total crack length of 25 inches. This value exceeds the critical crack length for-fracture which was computed to be 15 inches.

A summary of the critical flaw size for each weld flaw while maintaining the same aspect ratio is summarized in Table 7-1. These results indicate a factor of safety on flaw size between 10 to 15 which satisfies the flaw evaluation acceptance criteria of Eqs. 4-3 and 4-5. The calculated failure loci for part-through and part-circumferential flaws are shown in Figure 7-4.

The detected flaws are remote from the failure curves.

The calculated critical stress to cause pipe failure for each of the given flaw sizes is summarized in Table 7-2. The ratio of stress to cause a plastic collapse to the applied stress gives a safety factor between 2.5 and 3.0 for the weld defects. It should be noted that these safety factors were computed based on the limit stress, a.

The Code safety factors in B31.1 or Section III are based ona uts which is greater than a assumed herein. When this computed failure load for the flawed pipe model is compared to an analysis for a 24-inch unflawed pipe, the reduction in strength against plastic collapse is only 0.3%.to 5%. These results are shown in Table 7-3.

t F.

F-r r u-p

-=-

-7 1.77

........ 7777...

7.f7 7-7 a 1 -L;--

a 7-7 a

T-m 7.:

7.

7 T L

-777.

r 7

-_ 7 1: 7:

M..

q:

1::.

.11:

1

::::i

• ,I:::.

7=

a q :.

r 7 ---

7=

d:

7.

.7

T: ------ 177 d...

q Z4 a

7...

7

-7 7

-'----777777r777777777

77. 7-7:

LL:. I

......:T :

7 J:

-77 L

17 717...........

L

-vol 7=77.

.77 j

_7 7-Z C-17 c,4 L

q--

ZL I.

+

1; F~~~ji j

j'I ri

' ~

i:;'..

w

-7 II 7 7 7 7 -,

..1

....7 T F

'F-

_77F

'.=,.~

F

_ II dbTo 4I tlF' K F.

4 I__ '1 F' I 2

F...

IF 7

TF

.1 71

~i

.i 11

7-6 TABLE 7-1.

SUMMARY

OF CRITICAL FLAW SIZE RESULTS FROM LIMIT LOAD ANALYSIS Weld Flaw Type Applied Stress Critical Flaw Size Number and Location m/a b/Gk a./t ac/

ac/a 1-3 Slag Inclusions 0.147 0.395 1.0 0.33 15.4 (subsurface) 1-5 Slag Inclusions 0.154 0.314 1.0 0.36 15.4 L

(subsurface) 2-2 Slag Inclusions 0.157 0.305 0.72 0.70 11.1 (subsurface)1 Slag Inclusions 0.157 Q.305 0.87 0.42 13.4 (subsurface)

Lack-of-Penetration 0.157 0.305 0.94 0.38 12.9 OID surface) 2-3 Slag Inclusions 0.147 0.396 0.65 (subsurface) 1.0 10.0 6-2 Slag Line 0.156 0.385 0.65 1.0 10.0 (subsurface)

Acceptance Criterion:

a /a > 10 c.

'7

'7 A k

+T1_t+ 4_4_

1 1111

+H 1-

-4 M.

'7HE -TTET t

T 441+

Jtt

,r J-1

.1 T

T T i 4:

I i

1 4 'rT T, 7

-I :I-T IT ri 4 T_;-

IrTM.-

Tm

, PH

4. 4.

t 4

.4 1!- 71 i f

+

tt+

444+

j, Lt i j

_T S,

-T q

4+4

+

4 r

t Ti J i + _Tjfl I

flT

_1 iT C14 i I -I t 41:i CY) 14 F7 rl I

TIT w

-rt 7k

-I-t TTHT J

H-1 Lp

-4 4t

+

4-

+ +/-i+/-+/-

-1

_T' 1 T

1-

.7 44 T

-tff 4

I 4

'A A

4.

T-! 1 1,

T -1 T

a

+

11 I-T-I

+

T 7r H4

_7 77 77 -

T-rT-r #

P+

T 4 4

-T _44TT 4

7 7 T

4111 HIJ I i. t i 111-11 t

t

-I I

i d kf

+

4. 7 1 -41

--T r w

4--

+

-1 1 1

7 77-j.,

7!

_Z TT of jo T4+

ili 4+4 t f 14 li IT

+

t

+

TO lu TTi.

j 01711

-1,it ;H -

I, I T

I T

! I:T _14,LT ;_

144, 1-1 wV)

+4/)_

i 4 1+

0 d

ji 4i iA I T -,i i

i T

1.7 T:

I

-I-H it Lt I

10111 j-1T A

1-, I.

i, 11 T 1-t 7 TT -it 4

t4.

+

77

_T-1-i.

1-4 t:

t:t:t:t:t:

-I T: I I JT-'.

t t:

1-T q

t:

t --

i 1

,

!1H T,

- T 4 fi-I K--"-

W 1-t T _j:

t4+

4 1 PT

-r-t 1

-1 1: L

-.1 Tt-I 1 7 1-4

+

T 11 4

4,T I-I it 4 t

TI!.

1 4 T Tl.

J=_r p,

.14 t-i 1 ml

+ i j !J J

-TT-

q. T't J

_7 7

T, 1 t I

-tlft i

ifit ITT t q

i l

t I

7-8 TABLE 7-2

SUMMARY

OF LIMIT LOAD RESULTS FOR CRITICAL STRESS Weld Flaw Penetration Crack Half Angle Failure Stress*

Number Type (a/t or 2a/t)

(a/

Ratio, Gc/o 1-3 Subsurface 0.065 0.014 2.63 1-5 Subsurface 0.065 0.014 3.01 2-2 Subsurface 0.065 0.066 3.01 Subsurface 0.065 0.027 3.03 Surface 0.073 0.027 3.03 2-3 Subsurface 0.065 0.504 2.51 6-2 Subsurface 0.065 0.106 2.60

• 0= mo+b and 7c is defined as the total applied stress to cause failure.

Acceptance Criterion:

y c /a > r c

NOTE:

Failure Stress Ratio ac/a < /1 for unflawed pipe.

7-9 TABLE 7-3.

RATIO OF CALCULATED PLASTIC COLLAPSE LOAD FOR FLAWED PIPE TO UNFLAWED PIPE Reduction Weld Critical/Appl'ied Stress Ratio in Number Flaw Type o (flawed)/op(unflawed)

Strength 1-3 Slag Inclusions 0.997 0.3%

1-5 Slag Inclusions 0.997 0.3%

2-2 Slag Inclusions 0.989

1.

1%

Slag Inclusions 0.994 0.6%

Lack-of-Penetration 0.994 0.6%

2-3 Slag Inclusions 0.950 5.0%

6-2 Slag Line 0.984 1.6%

m-

8-1

8.0 CONCLUSION

S Based on this investigation, the following conclusions have been 6w reached:

(1) Although the flaw indications have been detected in five welds in the main steam piping system, these defects are not significant with regard to the static strength of the piping (i.e., fracture under monotonic loading). This conclusion.is based on a review of inspection acceptance standards, and a flaw evaluation of the inspection data.

(2) In the Summer 1975 Addenda to the 1974 Edition of Section XI of the ASME Boiler and Pressure Vessel Code, there are no acceptance standards given for defects in piping weldments.

(3) However, the Summer 1978 Addenda to the 1977 Edition of Section XI contain applicable piping inspection standards and all flaw indications presented herein are acceptable to this version of the Code.

(4) The indications have been evaluated using linear elastic fracture mechanics (LEFM) consistent with the flaw evaluation procedure of Appendix A in Section XI of the ASME Code.

This method predicts very large critical flaws, and the required stresses to cause failure must be much greater (10 times greater) than the stresses supplied to us by Bechtel.

Hence the failure mechanism will be controlled by a limit load type failure.

(5) An elastic-plastic fracture analysis using crack opening displacement (COD) concepts was also performed to determine Ithe parameters controlling failure. This method also indicated that loads much greater than those which have been provided to us would be required to cause failure.

8-2 (6)

Using limit load analysis methods,a failure locus was developed for combined membrane plus bending stresses.

Under these conditions of loading the flaw indications that we have analyzed are remote from failure locus.

These flaw indications would have to be scaled by a factor of about 10 and greater to reach the failure locus.

In this analysis a flow stress (a ) of 43.7 ksi was used which was the average of the yield stress and ultimate stress of the material at 540 0F.

(7) The above analysis methods indicate that leak-before-break is assured, and once the analysis has proceeded to the extent that the flaw has broken through the pipe surface the flaw is then analyzed as a through-thickness defect.

It indicates that for such flaws to become critical (i.e., for a break to occur) the extent of the flaw would have to be greater than 70 0 around the pipe circumference.

A-1 CONTROLLED DOCUMENTS REFERENCE LIST (CD)

1.

SCE SONGS 1 Radiographs, Line 1-24-EG, Weld 1-24-3, May 15, 1980.

2.

SCE SONGS 1 Radiographs, Line 1-24-EG, Weld 1-24-5, May 15, 1980.

3.

SCE SONGS 1 Radiographs, Line 6-24-EG, Weld 6-24-2, May 28, 1980.

4.

SCE SONGS 1 Radiographs, Line 2-24-EG, Weld 2-24-3, May 22, 1980.

S5.

SCE SONGS 1 Radiographs, Line 393-10-EG, Weld 393-10-EG, May 22, 1980.

6.

SCE SONGS 1 Radiographs, Line 2-24-EG, Weld 2-24-2, May 22, 1980.

7.

Bechtel Welding Standard -

Procedure Specification P1-AT-Lh-f, Rev. 3, February 23, 1965.

8.

Bechtel Calculation Sheets done May 29, 1980 by Doug Freeland.

9.

Flaw Evaluation Data, Welds 1-24-3, 1-24-5, and 2-24-2, May 30, 1980.

10.

Flaw Evaluation Data, Welds 2-24-3 and 6-24-2, May 30, 1980..

11.

Letter Report to W. M. Frazier (Bechtel) from G. Egan (Aptech), June 9, 1980.

12.

Letter Report to G. R. Egan (Aptech) from Robert Zweigler (Bechtel),

June 11, 1980.

13.

Photographs of Radiographic Image Enhancement, June 18, 1980

14.

Mobil Inspection Service Ultrasonic Flaw Evaluation Data, June 23, 1980.

A-2 CONTROLLED DRAWINGS REFERENCE LIST (CDD)

1. Drawing No. SCE 2-1A, "Main Steam Lines -

Stm. Gens. to Pene's.,"

August 4, 1977.

2. Drawing No. 456487-0, "Isometric Line 1-24-EG, 2-24-EG, 6-24-EG and 7-24-EG; Line Loop from 1-J-1 Pen's," Rev. 0, September 9, 1977.

3. Drawing No. 334533-2, "Isometric Line 5-20-EG from Stm. Gen. E-1B to 6-24-EG, 7-24-EG," Rev. 2, August 15, 1978.

4. Drawing No. 456474-0, "Isometric Line No. 392-10-EG from FOV-457 through Pen. C-3C to Stm. Gen. E-lb, Rev. 0, September 9, 1977.

aM don