14.07.2013 Views

iter structural design criteria for in-vessel components (sdc-ic)

iter structural design criteria for in-vessel components (sdc-ic)

iter structural design criteria for in-vessel components (sdc-ic)

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

ITER G 74 MA 8 01-05-28 W0.2<br />

ITER STRUCTURAL DESIGN CRITERIA<br />

FOR IN-VESSEL COMPONENTS<br />

(SDC-IC)<br />

APPENDIX B<br />

GUIDELINES FOR ANALYSIS,<br />

IN-VESSEL COMPONENTS


ITER G 74 MA 8 01-05-28 W0.2<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page ii


ITER G 74 MA 8 01-05-28 W0.2<br />

TABLE OF CONTENTS<br />

BÊ1000 GENERAL ............................................................................................................................ 1<br />

B 1100 Introduction.......................................................................................................................................1<br />

B 1110 Purpose.........................................................................................................................................1<br />

B 1120 Organization.................................................................................................................................1<br />

B 1130 Related documents.......................................................................................................................1<br />

B2300 Cr<strong>iter</strong>ia levels....................................................................................................................................1<br />

BÊ2301 Load Comb<strong>in</strong>ations and Associated Cr<strong>iter</strong>ia Levels .............................................................1<br />

B2500 Stress def<strong>in</strong>ition and classif<strong>ic</strong>ation...................................................................................................2<br />

BÊ2501 Mechan<strong>ic</strong>al stress.........................................................................................................................2<br />

BÊ2502 General and local thermal stress <strong>in</strong> a shell..................................................................................2<br />

BÊ 2503 Swell<strong>in</strong>g <strong>in</strong>duced stress ..........................................................................................................2<br />

B 2510 Breakdown of stresses .................................................................................................................3<br />

BÊ2513 Membrane stress .....................................................................................................................4<br />

BÊ2514 Bend<strong>in</strong>g stress .........................................................................................................................5<br />

BÊ2520 Classif<strong>ic</strong>ation of stresses obta<strong>in</strong>ed by elast<strong>ic</strong> analysis................................................................6<br />

B 2521 Primary stress..........................................................................................................................6<br />

B 2521.1 Primary membrane stress ...............................................................................................7<br />

B 2525 Secondary stress......................................................................................................................7<br />

B 2526 Peak stress...............................................................................................................................7<br />

B 2540 Stress Intensities / Equivalent stresses......................................................................................10<br />

BÊ2540.1 Stress <strong>in</strong>tensity - Maximum shear stress theory (Tresca) ...........................................10<br />

BÊ2540.2 Stress <strong>in</strong>tensity - Octahedral shear stress theory (von Mises) ....................................10<br />

B 2541 Hydrostat<strong>ic</strong> stress ( sH ).......................................................................................................11<br />

B 2541.1 Triaxiality factor...........................................................................................................11<br />

B 2550 Stress <strong>in</strong>tensity ranges/Equivalent stress ranges..................................................................11<br />

BÊ2550.1 Stress <strong>in</strong>tensity range - Maximum shear stress theory................................................11<br />

BÊ2550.2 Stress Intensity range - Octahedral shear stress theory...............................................14<br />

B 2600 Stra<strong>in</strong> def<strong>in</strong>itions and classif<strong>ic</strong>ation ...............................................................................................16<br />

B 2620 Calculation of equivalent stra<strong>in</strong> ( e) .........................................................................................16<br />

BÊ2630 Calculation of the equivalent stra<strong>in</strong> range ( De )......................................................................16<br />

B 2700 Terms related to limit quantities.....................................................................................................17<br />

B 2750 Terms related to fatigue damage ...............................................................................................17<br />

B 2752 Fatigue usage fraction V.......................................................................................................17<br />

BÊ2752.1 Procedure <strong>for</strong> comb<strong>in</strong>ation of cycles ...........................................................................17<br />

B3000 Design rules <strong>for</strong> s<strong>in</strong>gle-layer homogeneous structures .................................................. 20<br />

B 3020 Methods of analysis ...................................................................................................................20<br />

B 3021 Test to determ<strong>in</strong>e if nonl<strong>in</strong>ear (f<strong>in</strong>ite de<strong>for</strong>mation) analysis is needed ..............................21<br />

B 3022 Negligible irradiation-<strong>in</strong>duced swell<strong>in</strong>g test........................................................................22<br />

B 3023 Elast<strong>ic</strong> Analysis....................................................................................................................23<br />

B 3024 Inelast<strong>ic</strong> analysis...................................................................................................................23<br />

B 3024.1 Simplified <strong>in</strong>elast<strong>ic</strong> analysis.........................................................................................23<br />

B 3024.1.1 Elast<strong>ic</strong>-irradiation-<strong>in</strong>duced-creep analysis..........................................................23<br />

B 3024.1.2 Neuber's rule ........................................................................................................26<br />

B 3024.1.3 Elast<strong>ic</strong> follow-up factor (r)..................................................................................27<br />

BÊ3024.2 Elasto-plast<strong>ic</strong> analysis of a structure subjected to a monoton<strong>ic</strong> load<strong>in</strong>g ....................29<br />

BÊ3024.2.1 Use of the tensile curve .......................................................................................30<br />

BÊ3024.2.2 Plast<strong>ic</strong>ity cr<strong>iter</strong>ion................................................................................................31<br />

BÊ3024.2.3 Flow rule ..............................................................................................................32<br />

BÊ3024.2.4 Harden<strong>in</strong>g rule .....................................................................................................33<br />

BÊ3024.3 Elasto-plast<strong>ic</strong> analysis of a structure subjected to cycl<strong>ic</strong> load<strong>in</strong>g...............................33<br />

B 3024.4 Elasto-visco-plast<strong>ic</strong> analysis of a structure subjected to cycl<strong>ic</strong> load<strong>in</strong>g.....................34<br />

BÊ3024.5 Limit analysis (collapse load) ......................................................................................34<br />

BÊ3025 Zones of calculation..............................................................................................................34<br />

BÊ3026 Comb<strong>in</strong>ation of analysis methods ........................................................................................35<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page iii


ITER G 74 MA 8 01-05-28 W0.2<br />

B3030 Appl<strong>ic</strong>able rules - Flow of analysis...........................................................................................35<br />

B3031 Master flow charts <strong>for</strong> satisfy<strong>in</strong>g <strong>design</strong> rules.....................................................................35<br />

B 3040 Rules <strong>for</strong> the prevention of excessive de<strong>for</strong>mation affect<strong>in</strong>g functional adequacy ................36<br />

BÊ3050 Negligible thermal creep test.....................................................................................................36<br />

B 3100 LOW TEMPERATURE RULES...................................................................................................37<br />

B 3101 Negligible irradiation-<strong>in</strong>duced creep test.............................................................................37<br />

B 3200 Rules <strong>for</strong> the prevention of M type damage ..................................................................................38<br />

B 3211 Immediate plast<strong>ic</strong> collapse and plast<strong>ic</strong> <strong>in</strong>stability ...............................................................38<br />

B 3211.1 Elast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> collapse and plast<strong>ic</strong> <strong>in</strong>stability)..........................38<br />

B 3211.1.1 Bend<strong>in</strong>g shape factor ...........................................................................................38<br />

B 3211.2 Elasto-plast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> collapse)...................................................39<br />

B 3212 Immediate plast<strong>ic</strong> flow localization .....................................................................................39<br />

B 3212.1 Elast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> flow localization)................................................39<br />

B 3212.2 Elasto-plast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> flow localization) ....................................40<br />

B 3213 Immediate local fracture due to exhaustion of ductility......................................................40<br />

B 3213.1 Elast<strong>ic</strong> analysis (Immediate local fracture due to exhaustion of ductility)...............40<br />

B 3213.2 Elasto-plast<strong>ic</strong> analysis (Immediate local fracture due to exhaustion of ductility) .....41<br />

BÊ3214 Fast fracture..........................................................................................................................41<br />

B 3214.1 Elast<strong>ic</strong> analysis (Fast fracture)....................................................................................41<br />

B 3214.2 Elasto-plast<strong>ic</strong> analysis (Fast fracture).........................................................................42<br />

B 3300 Rules <strong>for</strong> the prevention of C type damage (Levels A and C)......................................................42<br />

B 3310 Progressive de<strong>for</strong>mation or ratchet<strong>in</strong>g ......................................................................................42<br />

B 3311 Elast<strong>ic</strong> analysis (Progressive de<strong>for</strong>mation or ratchet<strong>in</strong>g)....................................................42<br />

B 3311.1 3Sm rule........................................................................................................................42<br />

B 3311.2 Bree-diagram rule.........................................................................................................42<br />

B 3312 Elasto-plast<strong>ic</strong> analysis (Progressive de<strong>for</strong>mation or ratchet<strong>in</strong>g).........................................43<br />

B 3320 Time-<strong>in</strong>dependent fatigue..........................................................................................................44<br />

B 3322 Limits on fatigue damage .....................................................................................................44<br />

B 3322.1 Calculation of the fatigue usage fraction: V( De).....................................................44<br />

B 3322.2 Estimation of irradiation effects on fatigue usage fraction.........................................44<br />

BÊ3322.2.1 Calculation of fatigue usage fraction - general...................................................44<br />

BÊ3322.2.2 Procedure <strong>for</strong> estimat<strong>in</strong>g the fatigue curve.........................................................45<br />

B 3323. Calculation of equivalent stra<strong>in</strong> range De .........................................................................46<br />

B 3323.1 Elast<strong>ic</strong> analysis (Time-<strong>in</strong>dependent fatigue) ..............................................................46<br />

B 3323.1.1 Elast<strong>ic</strong> stra<strong>in</strong> range...............................................................................................46<br />

B 3323.1.2 Corrections <strong>for</strong> effects of plast<strong>ic</strong>ity.....................................................................47<br />

B 3323.1.3 Comb<strong>in</strong><strong>in</strong>g Components......................................................................................51<br />

B 3400 Rules <strong>for</strong> the prevention of buckl<strong>in</strong>g..............................................................................................52<br />

B 3420 Buckl<strong>in</strong>g limits...........................................................................................................................52<br />

B 3421 Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability under monoton<strong>ic</strong> load<strong>in</strong>g ...........................................52<br />

B 3421.1 Elast<strong>ic</strong> analysis (Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability )...............................................52<br />

B 3421.1.1 Load-controlled buckl<strong>in</strong>g limits..........................................................................52<br />

B 3421.1.2 Buckl<strong>in</strong>g diagrams...............................................................................................55<br />

B 3421.1.2 Stra<strong>in</strong>-controlled buckl<strong>in</strong>g limits ........................................................................60<br />

B 3421.2 Elasto-plast<strong>ic</strong> analysis (Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability).....................................60<br />

B 3500 High Temperature Rules.................................................................................................................61<br />

BÊ3800 Design Rules <strong>for</strong> bolted jo<strong>in</strong>ts........................................................................................................62<br />

BÊ3810 Methods of analysis ...................................................................................................................62<br />

BÊ3811 Elast<strong>ic</strong> analysis......................................................................................................................62<br />

B 3812 Simplified elast<strong>ic</strong> analysis ....................................................................................................63<br />

B 3812.1 Simplified uniaxial analysis.........................................................................................64<br />

B 3812.1.1 Effects of a temperature rise................................................................................65<br />

B 3812.1.2 Effects of an external load applied to the bolted assembly................................66<br />

B 3812.1.3 Effects of an external moment applied to the bolted assembly .........................67<br />

BÊ3813 Elasto-plast<strong>ic</strong> analysis...........................................................................................................69<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page iv


ITER G 74 MA 8 01-05-28 W0.2<br />

BÊ1000 GENERAL<br />

B 1100 INTRODUCTION<br />

This document provides guidel<strong>in</strong>es and procedures <strong>for</strong> analysis wh<strong>ic</strong>h should be used <strong>in</strong><br />

satisfy<strong>in</strong>g the rules presented <strong>in</strong> SDC-IC.<br />

B 1110 Purpose<br />

The purpose of appendix B is to provide analysis methods and guidel<strong>in</strong>es compatible with the<br />

<strong>design</strong> rules of SDC-IC. The primary <strong>in</strong>tent of this document is to facilitate the <strong>design</strong>er /<br />

analyst's job by provid<strong>in</strong>g widely accepted analysis guidel<strong>in</strong>es and <strong>in</strong>terpretation of the<br />

<strong>design</strong> rules.<br />

B 1120 Organization<br />

Guidel<strong>in</strong>es are presented <strong>in</strong> sections wh<strong>ic</strong>h are numbered, where appropriate, on a one-to-one<br />

correspondence with those <strong>in</strong> the SDC-IC. Cross-reference and redundancy are kept to a<br />

m<strong>in</strong>imum.<br />

General guidel<strong>in</strong>es <strong>for</strong> elast<strong>ic</strong> analysis and <strong>in</strong>elast<strong>ic</strong> analysis are presented <strong>in</strong> sections B 3023<br />

and B 3024, respectively.<br />

B 1130 Related documents<br />

Although this appendix is <strong>in</strong>tended to be self-conta<strong>in</strong>ed, references to SDC-IC, wh<strong>ic</strong>h conta<strong>in</strong><br />

the <strong>design</strong> rules, are frequently made. Appendix C conta<strong>in</strong>s the rationale or justif<strong>ic</strong>ations <strong>for</strong><br />

the <strong>design</strong> rules and may provide some <strong>in</strong>sights <strong>in</strong>to the <strong>in</strong>terpretation of the analysis methods<br />

presented here. Appendix A provides the materials <strong>design</strong> limit data used <strong>in</strong> the SDC-IC.<br />

Irradiation-<strong>in</strong>duced creep and stress-free swell<strong>in</strong>g properties are given <strong>in</strong> MAR 1 and MPH-<br />

IV 2 .<br />

B2300 CRITERIA LEVELS<br />

BÊ2301 Load Comb<strong>in</strong>ations and Associated Cr<strong>iter</strong>ia Levels<br />

Load<strong>in</strong>g category and damage limits specified <strong>in</strong> the DRG1 3 shall be used <strong>for</strong> <strong>structural</strong><br />

analysis. Specif<strong>ic</strong>ation of loads and their comb<strong>in</strong>ation that shall be used <strong>for</strong> analysis are given<br />

<strong>in</strong> LS document 4 . Relationship between load<strong>in</strong>g categories and cr<strong>iter</strong>ia levels are given <strong>in</strong> IC<br />

2200.<br />

1 G 74 MA 10, Materials Assessment Report<br />

2 G 74 MA 9, Materials Properties Handbook <strong>for</strong> In-<strong>vessel</strong> Components<br />

3 G A0 GDRD 2, Design Requirements and Guidel<strong>in</strong>es, Level 1<br />

4 G A0 MA 1, Load Specif<strong>ic</strong>ation and Comb<strong>in</strong>ation. Annex to DRG1.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 1


ITER G 74 MA 8 01-05-28 W0.2<br />

Neutron irradiation effects (e.g. embrittlement, swell<strong>in</strong>g, irradiation creep) must be <strong>in</strong>cluded<br />

<strong>in</strong> analysis.<br />

B2500 STRESS DEFINITION AND CLASSIFICATION<br />

BÊ2501 Mechan<strong>ic</strong>al stress<br />

Mechan<strong>ic</strong>al stresses are stresses wh<strong>ic</strong>h result from the appl<strong>ic</strong>ation of mechan<strong>ic</strong>al loads such<br />

as <strong>in</strong>ternal pressure, weight, earthquakes and, where appl<strong>ic</strong>able, reactions of supports and<br />

other <strong>components</strong>.<br />

BÊ2502 General and local thermal stress <strong>in</strong> a shell<br />

Thermal stresses are self-equilibrated stresses result<strong>in</strong>g from a non-uni<strong>for</strong>m spatial<br />

distribution of temperature or from the presence of materials with different thermal expansion<br />

coeff<strong>ic</strong>ients. For the purposes of apply<strong>in</strong>g stress cr<strong>iter</strong>ia, two types of thermal stresses are<br />

recognized: general thermal stresses and local thermal stresses, as def<strong>in</strong>ed <strong>in</strong> the follow<strong>in</strong>g.<br />

If a small portion of the shell is considered, that is, a portion whose dimensions<br />

normal to the median surface does not exceed a few th<strong>ic</strong>knesses, general thermal<br />

stresses entail a general de<strong>for</strong>mation (or stra<strong>in</strong>) of this portion; local thermal stresses<br />

do not <strong>in</strong>duce such general de<strong>for</strong>mation (or stra<strong>in</strong>).<br />

Examples of general thermal stress are<br />

- stress produced by axial temperature distribution <strong>in</strong> a cyl<strong>in</strong>dr<strong>ic</strong>al shell,<br />

- stress produced by temperature difference between a nozzle and the shell to wh<strong>ic</strong>h<br />

it is attached,<br />

- equivalent l<strong>in</strong>ear stress (IC 2515) distribution produced by a radial temperature<br />

distribution <strong>in</strong> a cyl<strong>in</strong>dr<strong>ic</strong>al shell,<br />

- equivalent l<strong>in</strong>ear stress (IC 2515) distribution produced by a through-th<strong>ic</strong>kness<br />

temperature distribution <strong>in</strong> a constra<strong>in</strong>ed flat plate.<br />

Examples of local thermal stress are<br />

- the stress <strong>in</strong> a small hot spot,<br />

- the non-l<strong>in</strong>early distributed (IC 2516) thermal stress, i.e., the difference between<br />

actual stress and equivalent l<strong>in</strong>ear stress,<br />

- thermal stress <strong>in</strong> a cladd<strong>in</strong>g material wh<strong>ic</strong>h has a different thermal expansion<br />

coeff<strong>ic</strong>ient than that of the base metal.<br />

BÊ 2503 Swell<strong>in</strong>g <strong>in</strong>duced stress<br />

These are self-equilibrated stresses result<strong>in</strong>g from a non-uni<strong>for</strong>m spatial distribution of<br />

fluence or from the presence of materials with different swell<strong>in</strong>g laws. As with thermal<br />

stresses (B2512), general and local swell<strong>in</strong>g stresses can be dist<strong>in</strong>guished. Swell<strong>in</strong>g-<strong>in</strong>duced<br />

stresses are limited by relaxation due to irradiation-<strong>in</strong>duced creep (IC 2151). Swell<strong>in</strong>g can<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 2


ITER G 74 MA 8 01-05-28 W0.2<br />

also lead to large de<strong>for</strong>mations of the structure wh<strong>ic</strong>h might require that all analysis <strong>in</strong>clude<br />

the effects of f<strong>in</strong>ite de<strong>for</strong>mation on stresses and stra<strong>in</strong>s (B 3021). On the other hand, under<br />

certa<strong>in</strong> circumstances (B 3022) if the temperature and neutron damages are suff<strong>ic</strong>iently low,<br />

the effects of irradiation-<strong>in</strong>duced swell<strong>in</strong>g may be neglected. If the effects of irradiation<strong>in</strong>duced<br />

swell<strong>in</strong>g is not negligible, the swell<strong>in</strong>g-<strong>in</strong>duced stress, <strong>in</strong>clud<strong>in</strong>g the relax<strong>in</strong>g effects<br />

of irradiation-<strong>in</strong>duced creep, has to be taken <strong>in</strong>to consideration either by detailed <strong>in</strong>elast<strong>ic</strong><br />

(elasto-visco-plast<strong>ic</strong>) analysis or by elast<strong>ic</strong>-irradiation-<strong>in</strong>duced creep analysis (B 3024.1.1.1).<br />

B 2510 Breakdown of stresses<br />

The decomposition of stresses <strong>in</strong>to membrane, bend<strong>in</strong>g, and peak categories is used <strong>for</strong><br />

elast<strong>ic</strong> analysis of shell and beam-like structures because, <strong>for</strong> these types of structures,<br />

different allowable stresses may be determ<strong>in</strong>ed (by a simple limit analysis) <strong>for</strong> the different<br />

stress categories. Primary membrane stress is limited by Sm. For primary bend<strong>in</strong>g stresses, a<br />

bend<strong>in</strong>g shape factor K accounts <strong>for</strong> redistribution of stress due to plast<strong>ic</strong>ity. For peak<br />

stresses, wh<strong>ic</strong>h are stra<strong>in</strong> controlled, there is no limit <strong>for</strong> a ductile material apart from fatigue.<br />

The l<strong>in</strong>e <strong>in</strong>tegration method used <strong>in</strong> IC 2513 - 2514 to decompose the stress <strong>components</strong> is<br />

str<strong>ic</strong>tly appl<strong>ic</strong>able only to homogeneous shells. With some modif<strong>ic</strong>ation (see below) the<br />

concept may be applied to beam-like structures. However, the concept of stress<br />

decomposition cannot be generalized <strong>for</strong> a three-dimensional structure. The treatment of<br />

these structure types is discussed below.<br />

For a 3-D structure, depend<strong>in</strong>g on the structure and the load<strong>in</strong>g, it is possible that the<br />

analyst can use judgement to decompose the stresses and apply the rules <strong>for</strong> elast<strong>ic</strong> analysis<br />

directly. When this is not possible, the recommended approach is to compare the applied<br />

load<strong>in</strong>gs to the limit load<strong>in</strong>gs that would cause failure, and to ensure that the safety factor on<br />

the applied load<strong>in</strong>g is consistent with the safety factors implied <strong>in</strong> the SDC-IC rules. An<br />

elast<strong>ic</strong> stress analysis is generally <strong>in</strong>suff<strong>ic</strong>ient to determ<strong>in</strong>e the limit load. There<strong>for</strong>e, <strong>for</strong> a<br />

three-dimentional structure, it may be approporate to use the rules <strong>for</strong> elasto-plast<strong>ic</strong> analysis<br />

(ICÊ3211.2, ICÊ3212.2, IC 3213.2, and ICÊ3214.2). Alternatively, the assessment may be<br />

based either on experiment or a nonl<strong>in</strong>ear analysis of a representative structure.<br />

The rema<strong>in</strong>der of this section addresses stress decomposition <strong>in</strong> structures other than three<br />

dimensional.<br />

For s<strong>in</strong>gle layer, homogeneous shells, each separate component of a stress tensor def<strong>in</strong>ed<br />

along the support<strong>in</strong>g l<strong>in</strong>e segment can be decomposed <strong>in</strong>to its membrane and bend<strong>in</strong>g<br />

<strong>components</strong>. The rules <strong>for</strong> the decomposition of stresses <strong>in</strong> a s<strong>in</strong>gle-layer homogeneous shell<br />

are given <strong>in</strong> IC 2513 - 2514.<br />

For shell or beam-like structures other than a s<strong>in</strong>gle-layer homogeneous shell, the<br />

decomposition of a stress component <strong>in</strong>to its membrane and bend<strong>in</strong>g <strong>components</strong> is not<br />

always straight<strong>for</strong>ward, and a general rule <strong>for</strong> decomposition of stresses cannot be given.<br />

The decomposition would depend on how the structure is modeled and the type of stress<br />

analysis conducted, to derive the distribution of stresses through the th<strong>ic</strong>kness. In some<br />

cases, the decomposition can be better implemented by an <strong>in</strong>tegration over an area rather than<br />

along a support<strong>in</strong>g l<strong>in</strong>e segment.<br />

As a simple example, consider a th<strong>ic</strong>k walled tube runn<strong>in</strong>g along the x2 direction. If the tube<br />

is analyzed as a shell, then the membrane and bend<strong>in</strong>g stress vary with position on the shell,<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 3


ITER G 74 MA 8 01-05-28 W0.2<br />

and a l<strong>in</strong>e <strong>in</strong>tegration through the th<strong>ic</strong>kness is used to calculate the breakdown of stresses at<br />

any position. On the other hand, if the tube is analysed as a beam, then the membrane and<br />

bend<strong>in</strong>g stress apply to the cross-section as a whole, and an area <strong>in</strong>tegral over the total crosssectional<br />

area of the tube would be more appropriate.<br />

Once a determ<strong>in</strong>ation has been made as to wh<strong>ic</strong>h type of <strong>in</strong>tegration (l<strong>in</strong>e <strong>in</strong>tegral or area<br />

<strong>in</strong>tegral) is more appropriate, a membrane stress component can be def<strong>in</strong>ed as the average or<br />

mean value of that stress component along that l<strong>in</strong>e or area. The bend<strong>in</strong>g component of the<br />

stress is a l<strong>in</strong>early vary<strong>in</strong>g stress wh<strong>ic</strong>h can be def<strong>in</strong>ed <strong>in</strong> such a way that its moment about<br />

the centroid of the l<strong>in</strong>e segment or the area is the same as the moment of the total stress<br />

component m<strong>in</strong>us the membrane stress component about the centroid.<br />

BÊ2513 Membrane stress<br />

As an example of a structure other than a s<strong>in</strong>gle layer homogeneous shell, consider the case<br />

of a first wall <strong>design</strong> consist<strong>in</strong>g of two face plates of th<strong>ic</strong>knesses h1 and h2 separated by a<br />

coolant channel of height hc where the coolant channels runn<strong>in</strong>g along one direction (say x2<br />

direction) are separated by regularly spaced webs á(Figure B 2513-1). In general, such a<br />

structure is anisotrop<strong>ic</strong> but could be analyzed us<strong>in</strong>g various geometr<strong>ic</strong>al approximations.<br />

X 3<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 4<br />

X 1<br />

Figure BÊ2513-1: First wall cross-section<br />

(1) At an elementary level, when x3 variation of stresses <strong>in</strong> the x1 direction can be<br />

neglected, a sl<strong>ic</strong>e of the first wall consist<strong>in</strong>g of halves of two adjacent coolant channels<br />

together with their face plates may be analyzed as a one-dimensional stress analysis<br />

problem. The cross-section of <strong>in</strong>terest is like an I-beam and the stress of <strong>in</strong>terest is the<br />

normal stress s22. For this case the membrane stress ( s22 ) has to be def<strong>in</strong>ed as the<br />

m<br />

average stress over the whole cross-section, i.e., based on an area <strong>in</strong>tegral.<br />

1<br />

( s ) = ò s<br />

m A<br />

22 22<br />

A<br />

dA (1)<br />

where A is the area of cross-section.<br />

(2) An alternative approach, when x3 variation of stresses <strong>in</strong> both directions are nonnegligible,<br />

could be to analyze the structure as an isotrop<strong>ic</strong>, homogeneous, and multi-<br />

h 1<br />

h c<br />

h 2


ITER G 74 MA 8 01-05-28 W0.2<br />

layer plate consist<strong>in</strong>g of many such coolant channels but ignor<strong>in</strong>g the contribution of<br />

the channel webs to the stiffnesses of the plate. For some appl<strong>ic</strong>ations, e.g., satisfy<strong>in</strong>g<br />

the primary stress limits, this approximation is conservative. In this case, a support<strong>in</strong>g<br />

l<strong>in</strong>e segment can be def<strong>in</strong>ed at all po<strong>in</strong>ts of the plate but it will be discont<strong>in</strong>uous<br />

because it will pass through the coolant channels. For such a multi-layer homogeneous<br />

shell, the <strong>components</strong> (sij)m can be def<strong>in</strong>ed by the follow<strong>in</strong>g equation:<br />

+ ht<br />

( ij ) = ( 1 )<br />

m ò-h<br />

ij 3<br />

s / h s dx<br />

(2)<br />

b<br />

where h = h1+ h2 = total solid th<strong>ic</strong>kness, exclud<strong>in</strong>g th<strong>ic</strong>kness of coolant<br />

channels, if any, and hb and ht are the distances of the extreme surfaces from<br />

the neutral plane.<br />

Axis x3 conta<strong>in</strong>s the support<strong>in</strong>g l<strong>in</strong>e segment of length hb + ht. The orig<strong>in</strong> of the x3 axis<br />

is taken at the centroid, i.e.,<br />

+ h<br />

t<br />

ò Jx ( 3) x3 dx3<br />

= 0<br />

(3)<br />

-h<br />

b<br />

where Jx ( 3)<br />

0<br />

= ì if x 3 is <strong>in</strong> the coolant channel<br />

í<br />

î1<br />

if x 3 is <strong>in</strong> the structure<br />

(3) A third alternative could be the same as the previous one but without ignor<strong>in</strong>g the<br />

stiffnesses of the channel webs <strong>in</strong> the x2 direction. so that the first wall can be modelled<br />

either as a multi-layer anisotrop<strong>ic</strong> plate or by us<strong>in</strong>g detailed f<strong>in</strong>ite-element analysis. In<br />

this case, the membrane <strong>components</strong> of the normal stresses can be derived by us<strong>in</strong>g Eq.<br />

2 <strong>for</strong> the s11 component and Eq. 1 <strong>for</strong> the s22 component. However, the membrane<br />

<strong>components</strong> <strong>for</strong> the shear stresses s12 and s21 def<strong>in</strong>ed by these equations are generally<br />

unequal.<br />

BÊ2514 Bend<strong>in</strong>g stress<br />

As <strong>in</strong> the case of membrane stress, <strong>for</strong> structures other than a s<strong>in</strong>gle-layer homogeneous shell,<br />

a general rule <strong>for</strong> decomposition of stresses <strong>in</strong>to their bend<strong>in</strong>g <strong>components</strong> cannot be given.<br />

The decomposition would depend on how the structure is modelled and the type of stress<br />

analysis conducted to derive the distribution of stresses through the th<strong>ic</strong>kness. Results <strong>for</strong> the<br />

three examples considered <strong>in</strong> section B 2513 are given below.<br />

(1) The bend<strong>in</strong>g stress tensor is given (as a function of x3) by the follow<strong>in</strong>g equation:<br />

x3<br />

s<br />

ù<br />

22 s<br />

b ij x3dA ëê I ûú ò<br />

(1)<br />

( ) = é<br />

A<br />

where I x dA<br />

= ò 3 2<br />

A<br />

and the orig<strong>in</strong> of the x3 axis is at the centroid of the area A.<br />

(2) The bend<strong>in</strong>g stress tensor is given (as a function of x3) by the follow<strong>in</strong>g equation:<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 5


ITER G 74 MA 8 01-05-28 W0.2<br />

( ) = é<br />

x3<br />

s<br />

ù +ht<br />

ij s<br />

b<br />

h<br />

ij x3 dx3<br />

ëê ûú ò<br />

(2)<br />

I - b<br />

+ h<br />

=<br />

-h<br />

t<br />

where I ò J( x3) x3 dx<br />

2<br />

b<br />

and J(x3) is def<strong>in</strong>ed <strong>in</strong> B 2513.<br />

Note: The bend<strong>in</strong>g shape factors <strong>for</strong> this case are given <strong>in</strong> Figure IC 3211-3<br />

3<br />

(3) In this case, the bend<strong>in</strong>g <strong>components</strong> of the normal stresses can be derived by us<strong>in</strong>g<br />

Eq.2 <strong>for</strong> the s11 component and Eq. 1 <strong>for</strong> the s22 component. However, the bend<strong>in</strong>g<br />

<strong>components</strong> <strong>for</strong> the shear stresses s12 and s21 def<strong>in</strong>ed by these equations are generally<br />

unequal.<br />

BÊ2520 Classif<strong>ic</strong>ation of stresses obta<strong>in</strong>ed by elast<strong>ic</strong> analysis<br />

General classif<strong>ic</strong>ations <strong>for</strong> stresses are given <strong>in</strong> IC 2520.<br />

B 2521 Primary stress<br />

The primary stress is def<strong>in</strong>ed as that portion of the total stress wh<strong>ic</strong>h is required to satisfy<br />

equilibrium with the applied load<strong>in</strong>g and wh<strong>ic</strong>h does not dim<strong>in</strong>ish after small scale permanent<br />

de<strong>for</strong>mation. Small scale permanent de<strong>for</strong>mation is taken to mean de<strong>for</strong>mation wh<strong>ic</strong>h results<br />

from plast<strong>ic</strong> stra<strong>in</strong>s that are of the same order of magnitude as the elast<strong>ic</strong> stra<strong>in</strong>s. If the<br />

stresses do not dim<strong>in</strong>ish after such small plast<strong>ic</strong> stra<strong>in</strong>s, then it is implied that the stresses<br />

cannot be relaxed by small de<strong>for</strong>mation and can lead to plast<strong>ic</strong> <strong>in</strong>stability and other damage.<br />

With<strong>in</strong> a structure, any stress field (e.g., the elast<strong>ic</strong> stress field or a lower bound stress field<br />

<strong>for</strong> limit analysis) wh<strong>ic</strong>h balances the volumetr<strong>ic</strong> <strong>for</strong>ces and the loads applied on the surface<br />

(mechan<strong>ic</strong>al loads: pressure, <strong>for</strong>ces, etc.) is an upper bound to the primary stress. This<br />

property is useful <strong>in</strong> pract<strong>ic</strong>e because the exact value of the primary stress is not always easy<br />

to determ<strong>in</strong>e. Thus, the upper bound will, more often than not, be used <strong>in</strong>stead of the true<br />

primary stress. The exact value of the primary stress may be obta<strong>in</strong>ed by tak<strong>in</strong>g the smallest<br />

stress field wh<strong>ic</strong>h balances the <strong>for</strong>ces, that is, that wh<strong>ic</strong>h leads to the lowest value of the<br />

maximum stress <strong>in</strong>tensity <strong>in</strong> the structure.<br />

Example: For a clamped edge beam with a uni<strong>for</strong>mly distributed load (w), from elast<strong>ic</strong><br />

analysis, an upper bound to primary bend<strong>in</strong>g stress (Pb) corresponds to the edge bend<strong>in</strong>g<br />

moment of wL 2 /12, where L is the span. However, a better (i.e., lower) upper bound to the<br />

primary bend<strong>in</strong>g stress, wh<strong>ic</strong>h can be obta<strong>in</strong>ed from limit analysis, corresponds to a bend<strong>in</strong>g<br />

moment of wL 2 /16.<br />

Generally, mechan<strong>ic</strong>al stresses (B2501) produced by mechan<strong>ic</strong>al loads are classified as<br />

primary.<br />

Note: If the elast<strong>ic</strong> stress analysis is conducted us<strong>in</strong>g a f<strong>in</strong>ite-element technique, the primary<br />

membrane and bend<strong>in</strong>g stresses at any section have to be determ<strong>in</strong>ed us<strong>in</strong>g the procedures<br />

given <strong>in</strong> IC 2513 and IC 2514, respectively (also, see sections B 2513 and B 2514).<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 6


ITER G 74 MA 8 01-05-28 W0.2<br />

B 2521.1 Primary membrane stress<br />

The def<strong>in</strong>ition of membrane stress is straight<strong>for</strong>ward <strong>for</strong> simple geometries, such as a solid<br />

shell or a section with several parallel solid shells, as def<strong>in</strong>ed <strong>in</strong> IC 2513. However, <strong>for</strong> more<br />

complex-shaped cross-section, the def<strong>in</strong>ition depends on how the geometry is modelled and<br />

analysed (see section B 2513).<br />

B 2525 Secondary stress<br />

Stresses aris<strong>in</strong>g from differential thermal expansion and swell<strong>in</strong>g, be<strong>in</strong>g de<strong>for</strong>mationcontrolled,<br />

are classified as secondary, except where the possibility of a large elast<strong>ic</strong> follow<br />

up (IC 2161) exists, <strong>in</strong> wh<strong>ic</strong>h case they should be classified as primary. General thermal (or<br />

constra<strong>in</strong>ed swell<strong>in</strong>g, B 2503) stresses (IC 2502) are generally classified as secondary.<br />

Elasto-plast<strong>ic</strong> analysis (B 3024) can be carried out to help dist<strong>in</strong>guish between primary and<br />

secondary stresses by apply<strong>in</strong>g the follow<strong>in</strong>g pr<strong>in</strong>ciple:<br />

"Any stress can be categorized as secondary, if it is likely to be redistributed <strong>in</strong> compliance<br />

with the signif<strong>ic</strong>ant stra<strong>in</strong> cr<strong>iter</strong>ia given <strong>in</strong> IC 3312."<br />

B 2526 Peak stress<br />

Peak stress is that <strong>in</strong>crement of stress wh<strong>ic</strong>h is additive to the primary and secondary stresses<br />

by reason of local discont<strong>in</strong>uities (IC 3024) or local thermal (or constra<strong>in</strong>ed swell<strong>in</strong>g, B<br />

2503) stress (B 2502) <strong>in</strong>clud<strong>in</strong>g the effects of, if any, stress concentrations. In a ductile<br />

material, peak stresses are objectionable only as a source of fatigue. However, <strong>in</strong> a lowductility<br />

material (e.g., irradiated sta<strong>in</strong>less steels), peak stress could also lead to local<br />

crack<strong>in</strong>g by exhaustion of ductility, and has to be guarded aga<strong>in</strong>st.<br />

Table B 2520-1 provides some guidance to the <strong>design</strong>er about stress classif<strong>ic</strong>ation. However,<br />

the <strong>design</strong>er is ultimately responsible <strong>for</strong> specify<strong>in</strong>g the f<strong>in</strong>al classif<strong>ic</strong>ation based on<br />

assessment of the specif<strong>ic</strong> <strong>design</strong> situation.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 7


ITER G 74 MA 8 01-05-28 W0.2<br />

Table B 2520-1: Stress classif<strong>ic</strong>ation <strong>for</strong> <strong>in</strong>-<strong>vessel</strong>s <strong>components</strong><br />

Location Orig<strong>in</strong> of stress Type of stress Classif<strong>ic</strong>ation Remarks<br />

Everywhere<br />

except at<br />

discont<strong>in</strong>uities<br />

Corners where<br />

two flats meet<br />

Internal/external<br />

pressure<br />

Thermal gradient<br />

through th<strong>ic</strong>kness<br />

Axial or<br />

circumferential<br />

thermal gradient<br />

Axial,<br />

circumferential<br />

or throughth<strong>ic</strong>kness<br />

swell<strong>in</strong>g gradient<br />

General<br />

membrane<br />

Bend<strong>in</strong>g<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 8<br />

Pm<br />

Pb<br />

Averaged across<br />

section<br />

L<strong>in</strong>ear variation<br />

across th<strong>ic</strong>kness<br />

Bend<strong>in</strong>g Q Thermal stress,<br />

de<strong>for</strong>mation<br />

controlled<br />

Membrane/<br />

Bend<strong>in</strong>g<br />

Membrane/<br />

Bend<strong>in</strong>g<br />

Internal pressure General<br />

Membrane<br />

Differential<br />

Swell<strong>in</strong>g/<br />

Temperature<br />

L<strong>in</strong>earized stress<br />

from bend<strong>in</strong>g<br />

moment<br />

Maximum stress<br />

<strong>in</strong>clud<strong>in</strong>g stress<br />

concentration<br />

Membrane<br />

L<strong>in</strong>earized<br />

component of<br />

bend<strong>in</strong>g<br />

Maximum stress<br />

<strong>in</strong>clud<strong>in</strong>g stress<br />

concentration<br />

Q De<strong>for</strong>mation<br />

controlled<br />

Q De<strong>for</strong>mation<br />

controlled<br />

Pm<br />

Pb<br />

Peak (F)<br />

Q<br />

Q<br />

Peak (F)<br />

Averaged across<br />

section<br />

l<strong>in</strong>ear variation<br />

across th<strong>ic</strong>kness<br />

Concentrated<br />

De<strong>for</strong>mation<br />

controlled<br />

De<strong>for</strong>mation<br />

controlled<br />

Concentrated


ITER G 74 MA 8 01-05-28 W0.2<br />

Table B 2520-1: Stress classif<strong>ic</strong>ation <strong>for</strong> <strong>in</strong>-<strong>vessel</strong>s <strong>components</strong> (Cont'd)<br />

Location Orig<strong>in</strong> of stress Type of stress Classif<strong>ic</strong>ation Remarks<br />

Junction with<br />

end plug/cap or<br />

near<br />

attachments<br />

Internal pressure Membrane<br />

Channel Channel to<br />

channel<br />

<strong>in</strong>teraction due to<br />

temperature<br />

difference across<br />

flats (bow<strong>in</strong>g)<br />

Fillets between<br />

flats<br />

Near holes,<br />

slits, etc.<br />

Channel to<br />

channel<br />

<strong>in</strong>teraction due to<br />

swell<strong>in</strong>g<br />

Internal/external<br />

pressure<br />

Channel to<br />

channel<br />

<strong>in</strong>teraction due to<br />

swell<strong>in</strong>g or<br />

temperature<br />

difference<br />

Stresses from any<br />

source<br />

Bend<strong>in</strong>g<br />

Bend<strong>in</strong>g<br />

Bend<strong>in</strong>g<br />

Membrane<br />

Bend<strong>in</strong>g<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 9<br />

PL<br />

Q<br />

Q<br />

Q<br />

PL<br />

Q<br />

Averaged across<br />

wall, acts at<br />

immediate<br />

v<strong>ic</strong><strong>in</strong>ity of<br />

junction<br />

Due to constra<strong>in</strong>t<br />

at end,<br />

de<strong>for</strong>mation<br />

controlled<br />

Thermal stress,<br />

de<strong>for</strong>mation<br />

controlled<br />

Swell<strong>in</strong>g <strong>in</strong>duced<br />

stress,<br />

de<strong>for</strong>mation<br />

controlled<br />

Averaged across<br />

wall, acts at<br />

immediate<br />

v<strong>ic</strong><strong>in</strong>ity<br />

Due to<br />

constra<strong>in</strong>t,<br />

de<strong>for</strong>mation<br />

controlled<br />

Bend<strong>in</strong>g Peak (F) Concentrated at<br />

fillet<br />

Local<br />

enhancement<br />

Peak (F) Concentrated


ITER G 74 MA 8 01-05-28 W0.2<br />

B 2540 Stress Intensities / Equivalent stresses<br />

BÊ2540.1 Stress <strong>in</strong>tensity - Maximum shear stress theory (Tresca)<br />

For a stress tensor with Cartesian <strong>components</strong> sij (i and j = 1, 2, 3) and pr<strong>in</strong>cipal <strong>components</strong><br />

s1, s2 and s3, the maximum shear stresses are related to the differences <strong>in</strong> the pr<strong>in</strong>cipal<br />

stresses as follows:<br />

2t12, max = s1 - s2<br />

2t23, max = s2 - s3<br />

2t31, max = s3 - s1<br />

The stress <strong>in</strong>tensity at any po<strong>in</strong>t is def<strong>in</strong>ed as tw<strong>ic</strong>e the largest absolute value of t12, max, t23,<br />

max, t31, max and is denoted by s<br />

( 1 2 2 3 3 1 )<br />

s = max s - s , s - s , s - s<br />

This <strong>for</strong>mula can be applied to the total stress tensor as well as to the stress tensor<br />

correspond<strong>in</strong>g to a stress category or a comb<strong>in</strong>ation of stress categories <strong>in</strong> the stress<br />

classif<strong>ic</strong>ation of ICÊ2520. When the stress <strong>in</strong>tensity of a comb<strong>in</strong>ation of several tensors is to<br />

be calculated, care must be taken to first sum the <strong>components</strong> of all tensors be<strong>for</strong>e calculat<strong>in</strong>g<br />

the stress <strong>in</strong>tensity. That is, one should calculate the stress <strong>in</strong>tensity of the sum of the tensors,<br />

not the sum of stress <strong>in</strong>tensities.<br />

BÊ2540.2 Stress <strong>in</strong>tensity - Octahedral shear stress theory (von Mises)<br />

For a stress tensor with Cartesian <strong>components</strong> sij (i and j = 1, 2, 3) and pr<strong>in</strong>cipal <strong>components</strong><br />

s1, s2 and s3, the stress <strong>in</strong>tensity at any po<strong>in</strong>t is equal to the value obta<strong>in</strong>ed by us<strong>in</strong>g any one<br />

of the follow<strong>in</strong>g three equivalent expressions:<br />

2<br />

2<br />

2<br />

{ ( 11 22 ) + ( 22 - 33 ) + ( 33 - 11)<br />

+ ( 12 + + ) }<br />

2 2<br />

23 31<br />

2<br />

s = 12. s - s s s s s 6 s s s<br />

2<br />

2<br />

{ ( 1 2)<br />

+ ( 2 - 3)<br />

+ ( 3 - 1)<br />

}<br />

s = 12.<br />

s - s s s s s<br />

{<br />

2<br />

s = s + s + s - s . s - s . s - s . s } 12<br />

1 2<br />

2<br />

3 2<br />

1 2 2 3 3 1<br />

The first expression is generally valid, whereas the follow<strong>in</strong>g two can be used only after the<br />

pr<strong>in</strong>cipal stresses have been determ<strong>in</strong>ed. As be<strong>for</strong>e (BÊ2540.1), these <strong>for</strong>mulae can also be<br />

applied to the stress tensor correspond<strong>in</strong>g to a comb<strong>in</strong>ation of stress categories, <strong>in</strong> wh<strong>ic</strong>h<br />

case, one should calculate the stress <strong>in</strong>tensity of the tensor sum.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 10<br />

2 12<br />

12


ITER G 74 MA 8 01-05-28 W0.2<br />

B 2541 Hydrostat<strong>ic</strong> stress ( s H )<br />

B 2541.1 Triaxiality factor<br />

The triaxiality factor (TF) is def<strong>in</strong>ed as the ratio between the hydrostat<strong>ic</strong> stress (IC 2541), sH,<br />

and the von Mises or octahedral shear stress (B 2540.2) normalized to unity <strong>for</strong> uniaxial<br />

load<strong>in</strong>g. Tests have shown that the ductility of a material can be signif<strong>ic</strong>antly reduced <strong>for</strong> a<br />

stress field with a high positive (i.e., tensile hydrostat<strong>ic</strong> stress) triaxiality factor, wh<strong>ic</strong>h<br />

typ<strong>ic</strong>ally occurs at the tip of a notch. There<strong>for</strong>e, all ductility-based <strong>design</strong> rules (IC 3212 and<br />

IC 3213) should account <strong>for</strong> the triaxiality effects by adjust<strong>in</strong>g the ductility (stra<strong>in</strong> limit)<br />

measured <strong>in</strong> uniaxial tests. For example, the S e rule <strong>for</strong> primary and secondary membrane<br />

stresses (ICÊ3212.1) reduces the uniaxial uni<strong>for</strong>m elongation by a factor of 2 to account <strong>for</strong><br />

biaxial membrane load<strong>in</strong>g. The S d rule <strong>for</strong> local fracture due to exhaustion of ductility<br />

(ICÊ3213.1) reduces the uniaxial true stra<strong>in</strong> of rupture by a factor = TF to account <strong>for</strong><br />

triaxiality <strong>in</strong> a notch.<br />

B 2550 Stress <strong>in</strong>tensity ranges/Equivalent stress ranges<br />

The rules <strong>for</strong> prevention of C-type damage require that calculated history of stress or stra<strong>in</strong> at<br />

any po<strong>in</strong>t <strong>in</strong> the structure be divided <strong>in</strong>to cycles. The procedure <strong>for</strong> deduc<strong>in</strong>g the cycles from<br />

the stress history is given <strong>in</strong> BÊ2752.1. Next, some of the rules require that the variation of<br />

the stress tensor dur<strong>in</strong>g a cycle be trans<strong>for</strong>med <strong>in</strong>to a scalar measure called the stress <strong>in</strong>tensity<br />

range. A simple procedure <strong>for</strong> calculat<strong>in</strong>g the stress <strong>in</strong>tensity range, when the <strong>components</strong> of<br />

the stress tensor vary directly <strong>in</strong> proportion to a scalar, is given <strong>in</strong> IC 2550.<br />

More general procedures, to be used when the variation of the stress tensor is not so simple,<br />

are given below <strong>for</strong> two common theories: maximum shear stress and octahedral shear stress.<br />

BÊ2550.1 Stress <strong>in</strong>tensity range - Maximum shear stress theory<br />

The general procedure <strong>for</strong> calculat<strong>in</strong>g the stress <strong>in</strong>tensity range us<strong>in</strong>g maximum shear stress<br />

(Tresca) theory is as follows:<br />

1) At each <strong>in</strong>stant (t) with<strong>in</strong> the cycle, calculate the <strong>components</strong> of the stress tensor<br />

s(t) at the po<strong>in</strong>t concerned.<br />

2) Calculate the tensor represent<strong>in</strong>g the stress difference s(t, t') <strong>for</strong> each pair of<br />

<strong>in</strong>stants (t) and (t') with<strong>in</strong> the cycle. The <strong>components</strong> of the tensor are equal to the<br />

difference between the <strong>components</strong> of tensors s(t) and s(t'):<br />

s(t, t') = s(t) - s(t')<br />

3) In accordance with ICÊ3224.4.2, calculate the stress <strong>in</strong>tensity s( tt ,©) of tensor<br />

s(t, t'). The stress <strong>in</strong>tensity range is thus the greatest of the absolute values of the<br />

follow<strong>in</strong>g quantities:<br />

S12(t, t') = s1(t, t') - s2(t, t')<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 11


ITER G 74 MA 8 01-05-28 W0.2<br />

S23(t, t') = s2(t, t') - s3(t, t')<br />

S31(t, t') = s3(t, t') - s1(t, t')<br />

Where <strong>in</strong>d<strong>ic</strong>es 1, 2, 3 denote the pr<strong>in</strong>cipal directions of the tensor s(t, t').<br />

Simplif<strong>ic</strong>ation:<br />

If the pr<strong>in</strong>cipal directions of the tensors s(t) rema<strong>in</strong> fixed with time, the pr<strong>in</strong>cipal<br />

directions of tensor s(t, t') co<strong>in</strong>cide with these directions <strong>for</strong> every pair of <strong>in</strong>stants<br />

(t) and (t') of the cycle. The pr<strong>in</strong>cipal stresses of s(t, t') can then be expressed by<br />

the follow<strong>in</strong>g equations:<br />

s1(t, t') = s1(t) - s1(t')<br />

s2(t, t') = s2(t) - s2(t')<br />

s3(t, t') = s3(t) - s3(t')<br />

4) For the cycle <strong>in</strong> question, the stress <strong>in</strong>tensity range is equal to the greatest of the<br />

quantities s tt , ©<br />

( ) calculated <strong>for</strong> every pair of <strong>in</strong>stants (t) and (t') of the cycle:<br />

[ ] = [ ( ) - ( ) ]<br />

Ds Max s t,© t Max s t s t©<br />

tt ,© tt ,©<br />

= ( )<br />

( ) ( )<br />

The search <strong>for</strong> the maximum value can be made easier if one of the two <strong>in</strong>stants<br />

def<strong>in</strong><strong>in</strong>g the cycle is fixed. If tA is taken as the fixed <strong>in</strong>stant, the stress range is<br />

equal to the greatest of the quantities s(t, tA) calculated <strong>for</strong> each <strong>in</strong>stant t of the<br />

cycle. The stress <strong>in</strong>tensity range is then given by:<br />

[ ]<br />

Ds Max s t s tA t<br />

= ( ) - ( )<br />

If the pr<strong>in</strong>cipal directions of the tensor s(t) rema<strong>in</strong> fixed with time, the search <strong>for</strong> the stress<br />

range us<strong>in</strong>g maximum shear stress theory can be carried out graph<strong>ic</strong>ally <strong>in</strong> the plane of the<br />

stress deviator s(t) (Figure BÊ2550-1). Indeed, the pr<strong>in</strong>cipal <strong>components</strong> of the stress deviator<br />

s(t) can be expressed as functions of the pr<strong>in</strong>cipal stresses s1(t), s2(t), s3(t) as follows:<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

s t s t s t s t s t<br />

1 1 1 2 3<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

s t s t s t s t s t<br />

2 2 2 2 3<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

s t s t s t s t s t<br />

3 3 1 2 3<br />

In the plane of the stress deviator (Figure BÊ2550-1), def<strong>in</strong>e three coplanar axes 1, 2 and 3<br />

with orig<strong>in</strong> at the po<strong>in</strong>t O and with unit vectors r r r<br />

i , j, and k <strong>in</strong>cl<strong>in</strong>ed at an angle 120¡ to each<br />

other. For a po<strong>in</strong>t M with Cartesian projection Ki on these axes, the follow<strong>in</strong>g equation<br />

OK = 32s ( t) i = 1to 3<br />

i i<br />

( )<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 12<br />

3<br />

3<br />

3


ITER G 74 MA 8 01-05-28 W0.2<br />

describes a closed plane curve dur<strong>in</strong>g the cycle <strong>in</strong> question. This closed plane curve is<br />

conta<strong>in</strong>ed with<strong>in</strong> an hexagon (shown by dotted l<strong>in</strong>es), the sides of wh<strong>ic</strong>h are parallel to the<br />

directions of axes 1, 2 and 3. The greatest of distances d23, d31 and d12, measured parallel<br />

to the directions of axes 1, 2 and 3 respectively between the sides of this hexagon divided by<br />

2 3 gives the stress <strong>in</strong>tensity range <strong>for</strong> the cycle concerned.<br />

( ) = ( )<br />

23 Ds Max d12, d23, d31<br />

Ds = ( 2) ( )<br />

12<br />

Max D , D , D<br />

r r r<br />

i, j, k : unit vectors<br />

1 2 3<br />

Figure BÊ2550-1: Stress <strong>in</strong>tensity range, maximum shear theory,<br />

fixed pr<strong>in</strong>ciple directions<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 13


ITER G 74 MA 8 01-05-28 W0.2<br />

BÊ2550.2 Stress Intensity range - Octahedral shear stress theory<br />

The general procedure <strong>for</strong> calculat<strong>in</strong>g the stress <strong>in</strong>tensity range us<strong>in</strong>g the octahedral shear<br />

stress (Von Mises) yield theory is as follows:<br />

1) At each <strong>in</strong>stant (t) of the cycle concerned, calculate the <strong>components</strong> of the stress<br />

tensor s(t) at the po<strong>in</strong>t <strong>in</strong> question.<br />

2) Calculate the tensor wh<strong>ic</strong>h represents the stress difference s(t, t') <strong>for</strong> each pair (t),<br />

(t') of the cycle. The <strong>components</strong> of the tensor s(t, t') are equal to the difference<br />

between the <strong>components</strong> of the tensors s(t) and s(t'):<br />

s(t, t') = s(t) - s(t')<br />

3) Us<strong>in</strong>g ICÊ3224.4.3, calculate the stress <strong>in</strong>tensity s( tt , © ) of tensor s(t, t'):<br />

{<br />

2<br />

[ 11 22 ] + [ 22 ( ) - 33(<br />

) ]<br />

( ) = ( ) - ( )<br />

s tt ,© 12.<br />

s tt ,© s tt ,© s tt ,© s tt ,©<br />

Simplif<strong>ic</strong>ation:<br />

[ s33 tt ,© s11<br />

tt ,© ]<br />

+ ( ) - ( )<br />

6 [ s12 tt ,© s tt ,© s tt ,© ]}<br />

2<br />

2<br />

23 31<br />

2<br />

+ ( ) + ( ) + ( )<br />

2<br />

If the pr<strong>in</strong>cipal directions of the stress tensors s(t) rema<strong>in</strong> fixed with time, the<br />

stress <strong>in</strong>tensity s tt , ©<br />

the stress deviator s(t):<br />

s t s t s t s t s t 3<br />

( ) is expressed as a function of the pr<strong>in</strong>cipal <strong>components</strong> of<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

1 1 1 2 3<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

s t s t s t s t s t<br />

2 2 1 2 3<br />

[ ]<br />

( ) = ( ) - ( ) + ( ) + ( )<br />

s t s t s t s t s t<br />

Then<br />

3 3 1 2 3<br />

2<br />

2<br />

( ) = ( ( ) - ( ) ) + ( ( ) - ( ) ) + ( ( ) - ( ) )<br />

s t,© t . s t s t© s t s t© s t s t©<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 14<br />

3<br />

3<br />

{ 2 2<br />

3 3 }<br />

32 1 1<br />

12<br />

2<br />

2 12<br />

4) For the cycle concerned, the stress <strong>in</strong>tensity range denoted by Ds is equal to the<br />

greatest of the quantities s( tt , © ) calculated <strong>for</strong> each pair of <strong>in</strong>stants (t) and (t') of<br />

the cycle:<br />

Ds Max<br />

tt ,©<br />

s<br />

[ ]<br />

= ( tt ,© )<br />

( )<br />

The search <strong>for</strong> the maximum value can be made easier if one of the two <strong>in</strong>stants<br />

def<strong>in</strong><strong>in</strong>g the cycle is fixed. If tA is taken as the fixed <strong>in</strong>stant, the stress range is


ITER G 74 MA 8 01-05-28 W0.2<br />

equal to the greatest of the quantities s(t, tA) calculated <strong>for</strong> each <strong>in</strong>stant t of the<br />

cycle. The stress <strong>in</strong>tensity range is then given by: :<br />

Ds Max<br />

t<br />

s<br />

[ ttA ]<br />

= ( , )<br />

If the pr<strong>in</strong>cipal directions associated with the stress tensors s(t) rema<strong>in</strong> fixed with time, the<br />

search <strong>for</strong> the range us<strong>in</strong>g the octahedral shear stress theory may be carried out graph<strong>ic</strong>ally<br />

(Figure BÊ2550-2). In the plane of the stress deviator, def<strong>in</strong>e three coplanar axes 1, 2 and 3<br />

with orig<strong>in</strong> at the po<strong>in</strong>t O and with unit vectors r r r<br />

i , j, and k <strong>in</strong>cl<strong>in</strong>ed at an angle 120¡ to each<br />

other. For a po<strong>in</strong>t M with Cartesian projection Ki on these axes, the follow<strong>in</strong>g equation<br />

OK = 3 2 S ( t) i = 1 to 3<br />

i i<br />

( )<br />

describes a closed plane curve dur<strong>in</strong>g the cycle concerned. The largest diameter of this curve<br />

divided by 2 3 gives the stress <strong>in</strong>tensity range <strong>for</strong> the cycle <strong>in</strong> question.<br />

r r r<br />

i, j, k : unit vectors<br />

Figure BÊ2550-2: Stress <strong>in</strong>tensity range, octahedral shear theory,<br />

fixed pr<strong>in</strong>ciple directions<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 15


ITER G 74 MA 8 01-05-28 W0.2<br />

B 2600 STRAIN DEFINITIONS AND CLASSIFICATION<br />

B 2620 Calculation of equivalent stra<strong>in</strong> ( e)<br />

For a stra<strong>in</strong> tensor with Cartesian <strong>components</strong> eij (i and j = 1, 2, 3), the equivalent stra<strong>in</strong> is<br />

given by<br />

2<br />

e = { [ e11 - e22 ] + [ e - e ] + e - e<br />

3<br />

2<br />

22 33 2<br />

.<br />

[ 12<br />

] }<br />

2 2<br />

23 31<br />

+ 6 e + e + e<br />

2 1 2<br />

[ ]<br />

33 11 2<br />

BÊ2630 Calculation of the equivalent stra<strong>in</strong> range ( De)<br />

The equivalent stra<strong>in</strong> range De at a po<strong>in</strong>t is used to compute the fatigue usage fraction. A<br />

simple procedure <strong>for</strong> calculat<strong>in</strong>g the equivalent stra<strong>in</strong> range, when the <strong>components</strong> of the<br />

stra<strong>in</strong> tensor vary directly <strong>in</strong> proportion to a scalar, is given <strong>in</strong> IC 2630.<br />

In general, when the <strong>components</strong> of the stra<strong>in</strong> tensor do not vary <strong>in</strong> direct proportion to a<br />

scalar, the stra<strong>in</strong> range is calculated <strong>in</strong> the follow<strong>in</strong>g way:<br />

1) At each <strong>in</strong>stant (t) of the cycle, calculate the <strong>components</strong> of the stra<strong>in</strong> tensor e(t) at<br />

the po<strong>in</strong>t exam<strong>in</strong>ed:<br />

e11(t); e22(t); e33(t); e12(t); e13(t); e31(t)<br />

2) Calculate the stra<strong>in</strong> range tensors e(t, t') <strong>for</strong> each pair of <strong>in</strong>stants (t) and (t') of the<br />

cycle. The <strong>components</strong> of tensor e(t, t') are equal to the difference between the<br />

<strong>components</strong> of tensors e(t) and e(t'):<br />

e(t, t') = e(t) - e(t')<br />

3) Calculate the equivalent scalar stra<strong>in</strong> range e tt , ©<br />

us<strong>in</strong>g one of the follow<strong>in</strong>g <strong>for</strong>mulae:<br />

{<br />

( ) between the states (t) and (t')<br />

2<br />

2<br />

e( tt ,© ) = . [ e11( tt ,© ) - e22 ( tt ,© ) ] + e22 ( tt ,© ) - e33(<br />

tt ,© )<br />

3<br />

[ e33 tt ,© e11<br />

tt ,© ]<br />

+ ( ) - ( )<br />

6 [ e12 tt ,© e tt ,© e tt ,© ] }<br />

2<br />

2<br />

23 31<br />

+ ( ) + ( ) + ( )<br />

2<br />

[ ]<br />

2 1 2<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 16<br />

2


ITER G 74 MA 8 01-05-28 W0.2<br />

{<br />

2<br />

2<br />

e( tt ,© ) = . [ e1( tt ,© ) - e2( tt ,© ) ] + e2( tt ,© ) - e3(<br />

tt ,© )<br />

3<br />

where<br />

[ e2 tt ,© e3<br />

tt ,© ] }<br />

+ ( ) - ( )<br />

2 12<br />

[ ]<br />

e11(t, t'), e12(t, t') are the <strong>components</strong> of tensor e(t, t'), and<br />

e1(t, t'), e2(t, t'), e3(t, t') are the pr<strong>in</strong>cipal <strong>components</strong> of this tensor.<br />

4) For the cycle exam<strong>in</strong>ed, the stra<strong>in</strong> range is equal to the greatest of the quantities<br />

e tt , ©<br />

( ) calculated <strong>for</strong> each pair of <strong>in</strong>stants (t) and (t') of the cycle:<br />

De Max<br />

tt ,©<br />

e<br />

[ ]<br />

= ( tt ,© )<br />

( )<br />

B 2700 TERMS RELATED TO LIMIT QUANTITIES<br />

B 2750 Terms related to fatigue damage<br />

B 2752 Fatigue usage fraction V<br />

BÊ2752.1 Procedure <strong>for</strong> comb<strong>in</strong>ation of cycles<br />

The follow<strong>in</strong>g procedure <strong>for</strong> comb<strong>in</strong>ation of cycles is recommended to ensure that a random<br />

sequence of cycles wh<strong>ic</strong>h could occur <strong>in</strong> the same time period are comb<strong>in</strong>ed <strong>in</strong> such a way<br />

that the calculated stra<strong>in</strong> ranges are a maximum, there<strong>for</strong>e conservative <strong>in</strong> a fatigue<br />

evaluation. For clarif<strong>ic</strong>ation, refer to the illustrated example below, wh<strong>ic</strong>h assumes that the<br />

stra<strong>in</strong> cycles are uniaxial.<br />

1. If two types of stra<strong>in</strong> cycles separately produce stra<strong>in</strong> ranges wh<strong>ic</strong>h are lower than<br />

those that would be produced by a s<strong>in</strong>gle type of cycle <strong>for</strong>med by a concatenation<br />

of these cycles, then the two cycles must then be comb<strong>in</strong>ed <strong>in</strong>to a s<strong>in</strong>gle cycle<br />

with the comb<strong>in</strong>ed stra<strong>in</strong> range.<br />

2. If two types of cycles are comb<strong>in</strong>ed <strong>in</strong> accordance with rule #1, and if n1 and n2<br />

are the orig<strong>in</strong>al numbers of cycles, and if n1 is less then n2, then the two types of<br />

cycle should be comb<strong>in</strong>ed <strong>for</strong> n1 cycles with a comb<strong>in</strong>ed range, leav<strong>in</strong>g (n2-n1)<br />

cycles of the second type.<br />

3. To preserve the total number of cycles, an additional number, n1, of cycles that<br />

<strong>in</strong>clude the <strong>in</strong>termediate maxima and m<strong>in</strong>ima must be taken <strong>in</strong>to account.<br />

4. The above process should be repeated, if necessary, until no more comb<strong>in</strong>ations<br />

are possible.<br />

The follow<strong>in</strong>g two figures illustrate these rules and expla<strong>in</strong> the log<strong>ic</strong>.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 17<br />

2


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure BÊ2752-1: Orig<strong>in</strong>al Cycles<br />

Example of rule #1. Observe that Cycle 1 has a stra<strong>in</strong> range of (50 + 10) = 60.<br />

Cycle 2 has a stra<strong>in</strong> range of (15+30) = 45. In accordance with rule #1 above,<br />

however, a larger stra<strong>in</strong> range (50 + 30) = 80 can be obta<strong>in</strong>ed by comb<strong>in</strong><strong>in</strong>g the<br />

positive half of cycle 1 with the negative half of cycle 2. This gives comb<strong>in</strong>ed<br />

cycle 1Õ shown below.<br />

Example of rule #2. Observe that the number of cycles of type 1 is 1000 and the<br />

number of type 2 is 10,000. In accordance with rule #2, the number of cycles of<br />

type 1Õ should be the smaller of these, or 1000. This leaves 10000 - 1000 - 9000<br />

cycles of type 2Õ, as shown below.<br />

Example of rule #3. The concatenation of orig<strong>in</strong>al cycles 1 and 2 gives a variation<br />

+50, -10, +15, -30. The comb<strong>in</strong>ation of these <strong>in</strong> accordance with the above,<br />

resulted <strong>in</strong> 1000 cycles with range +50 to -30. To account <strong>for</strong> the <strong>in</strong>termediate<br />

maxima and m<strong>in</strong>ima, 1000 additional cycles with range +15 to -10 must be<br />

<strong>in</strong>cluded. These are the type 3Õ cycles shown below.<br />

Figure BÊ2752-2: Comb<strong>in</strong>ed Cycles<br />

A generalization of the above procedure can be expressed as follows. In a history of random<br />

cycles, select the first composite cycle as one hav<strong>in</strong>g the maximum stra<strong>in</strong> and the m<strong>in</strong>imum<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 18


ITER G 74 MA 8 01-05-28 W0.2<br />

stra<strong>in</strong> dur<strong>in</strong>g the period. Remove that cycle from the history. Then, cont<strong>in</strong>ue select<strong>in</strong>g<br />

cycles, each time select<strong>in</strong>g the maximum and the m<strong>in</strong>imum of the ones that rema<strong>in</strong>.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 19


ITER G 74 MA 8 01-05-28 W0.2<br />

B3000 DESIGN RULES FOR SINGLE-LAYER<br />

HOMOGENEOUS STRUCTURES<br />

The appl<strong>ic</strong>able rules depend on whether or not the effects of thermal creep are negligible. If<br />

they are negligible, then the low temperature <strong>design</strong> rules (IC 3100) apply. If not, hightemperature<br />

rules (IC 3600) are appl<strong>ic</strong>able.<br />

B 3020 Methods of analysis<br />

Two alternative types of analyses are permissible - elast<strong>ic</strong> and <strong>in</strong>elast<strong>ic</strong>. Elast<strong>ic</strong> analysis<br />

should be the method of first cho<strong>ic</strong>e because it is much simpler. Inelast<strong>ic</strong> analysis is more<br />

complex and likely to encounter problems with convergence, accuracy, etc. Furthermore, the<br />

validation of <strong>in</strong>elast<strong>ic</strong> analysis methods, part<strong>ic</strong>ularly those deal<strong>in</strong>g with cycl<strong>ic</strong> stresses <strong>in</strong> a<br />

nonl<strong>in</strong>ear material, requires extensive theoret<strong>ic</strong>al and experimental work.<br />

The cr<strong>iter</strong>ia to be used <strong>in</strong> conjunction with elast<strong>ic</strong> analyses are purposely selected to be fairly<br />

conservative. Other methods of analysis could be pursued if elast<strong>ic</strong> analysis fails to<br />

demonstrate compliance with the cr<strong>iter</strong>ia.<br />

In general, <strong>in</strong>elast<strong>ic</strong> analyses methods should be adopted if the cr<strong>iter</strong>ia based on elast<strong>ic</strong><br />

analyses cannot be satisfied. These might <strong>in</strong>clude <strong>in</strong>elast<strong>ic</strong> f<strong>in</strong>ite element analysis and<br />

simplified <strong>in</strong>elast<strong>ic</strong> analysis methods. Guidel<strong>in</strong>es <strong>for</strong> satisfy<strong>in</strong>g limits <strong>for</strong> <strong>in</strong>elast<strong>ic</strong> analysis<br />

us<strong>in</strong>g simplified <strong>in</strong>elast<strong>ic</strong> analysis methods are given <strong>in</strong> this chapter. Other simplified<br />

analysis methods could be used provided they can be justified and shown to yield<br />

conservative results. In general, <strong>in</strong>elast<strong>ic</strong> f<strong>in</strong>ite element analyses should be conducted if the<br />

simplified <strong>in</strong>elast<strong>ic</strong> analysis methods fail to satisfy the cr<strong>iter</strong>ia <strong>for</strong> <strong>in</strong>elast<strong>ic</strong> analysis.<br />

In general, elast<strong>ic</strong> analysis is used to satisfy stress limits while <strong>in</strong>elast<strong>ic</strong> analysis is used to<br />

satisfy stra<strong>in</strong> limits. This is because elast<strong>ic</strong> analysis cannot calculate plast<strong>ic</strong> stra<strong>in</strong>s directly,<br />

render<strong>in</strong>g stress (and the equivalent elast<strong>ic</strong> stra<strong>in</strong>) as its only useful measure. Also, when an<br />

<strong>in</strong>elast<strong>ic</strong> analysis is per<strong>for</strong>med <strong>in</strong> the range of <strong>in</strong>terest, signif<strong>ic</strong>ant plast<strong>ic</strong> stra<strong>in</strong>s can occur<br />

with only small variation of stress, render<strong>in</strong>g stra<strong>in</strong> a more useful measure. In addition to the<br />

stress and stra<strong>in</strong> limits needed to ensure <strong>structural</strong> <strong>in</strong>tegrity, fast fracture limits, fatigue limits,<br />

buckl<strong>in</strong>g limit, and de<strong>for</strong>mation limits <strong>for</strong> functional adequacy must be satisfied <strong>for</strong> both<br />

types of analyses.<br />

Both elast<strong>ic</strong> and <strong>in</strong>elast<strong>ic</strong> analyses may be conducted on the assumption that the<br />

displacements and stra<strong>in</strong> are <strong>in</strong>f<strong>in</strong>itesimal (geometr<strong>ic</strong> l<strong>in</strong>earity). However, the analyst should<br />

verify that this assumption is reasonable, part<strong>ic</strong>ularly <strong>in</strong> the presence of irradiation-<strong>in</strong>duced<br />

swell<strong>in</strong>g and creep, because the <strong>design</strong> rules do not expl<strong>ic</strong>itly put a limit on these stra<strong>in</strong>s.<br />

Guidance <strong>for</strong> verify<strong>in</strong>g this assumption is given <strong>in</strong> B 3021. If the stra<strong>in</strong>s and displacements<br />

become signif<strong>ic</strong>ant, an elasto-visco-plast<strong>ic</strong> analysis, <strong>in</strong>clud<strong>in</strong>g f<strong>in</strong>ite de<strong>for</strong>mation effects may<br />

be needed.<br />

The <strong>design</strong> rules have been separated <strong>in</strong>to two classes - low-temperature rules (thermal creep<br />

effects can be neglected) and high-temperature rules (thermal creep effects cannot be<br />

neglected). To determ<strong>in</strong>e wh<strong>ic</strong>h rules are appl<strong>ic</strong>able, a test <strong>for</strong> negligible thermal creep<br />

(IC3050) has been <strong>in</strong>cluded.<br />

Stra<strong>in</strong> rate or time-dependent stra<strong>in</strong> (or creep) effects are considered expl<strong>ic</strong>itly <strong>in</strong> the hightemperature<br />

rules, with or without irradiation-<strong>in</strong>duced swell<strong>in</strong>g and creep. Generally, stra<strong>in</strong>-<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 20


ITER G 74 MA 8 01-05-28 W0.2<br />

rate or time-dependent stra<strong>in</strong> effects are ignored <strong>in</strong> the low-temperature rules. However, they<br />

may become important <strong>in</strong> stress analysis (by <strong>in</strong>troduc<strong>in</strong>g swell<strong>in</strong>g-<strong>in</strong>duced stress (B 2513)<br />

and by redistribut<strong>in</strong>g stresses and stra<strong>in</strong>s), wh<strong>ic</strong>h is required be<strong>for</strong>e the low-temperature rules<br />

can be applied, if signif<strong>ic</strong>ant amounts of irradiation-<strong>in</strong>duced swell<strong>in</strong>g and creep stra<strong>in</strong>s occur.<br />

There<strong>for</strong>e, be<strong>for</strong>e apply<strong>in</strong>g the low-temperature <strong>design</strong> rules, a determ<strong>in</strong>ation has to be made<br />

if the effects of irradiation-<strong>in</strong>duced swell<strong>in</strong>g (B 3022) and irradiation-<strong>in</strong>duced creep (B 3101)<br />

are negligible .<br />

B 3021 Test to determ<strong>in</strong>e if nonl<strong>in</strong>ear (f<strong>in</strong>ite de<strong>for</strong>mation) analysis<br />

is needed<br />

The follow<strong>in</strong>g procedure may be used to determ<strong>in</strong>e if nonl<strong>in</strong>ear (f<strong>in</strong>ite de<strong>for</strong>mation) analysis<br />

is needed.<br />

1) The total operat<strong>in</strong>g period of the component, throughout its life, and all load<strong>in</strong>gs,<br />

<strong>for</strong> wh<strong>ic</strong>h compliance with level A, C and D cr<strong>iter</strong>ia is required, are taken <strong>in</strong>to<br />

account.<br />

2) The total operat<strong>in</strong>g period is divided <strong>in</strong>to N <strong>in</strong>tervals of time. For each <strong>in</strong>terval i,<br />

of a duration ti, the th<strong>ic</strong>kness-averaged values of the maximum temperature<br />

reached Tmi, the mean neutron flux Fmi, the mean neutron fluence Ftmi (or<br />

displacement dose) and the correspond<strong>in</strong>g allowable primary membrane stress<br />

<strong>in</strong>tensity Smi (from Table A.MAT.5.1 of appendix A) are noted.<br />

3) To take full advantage of this rule, the <strong>in</strong>tervals of time must be chosen <strong>in</strong> such a<br />

way that the temperature and the flux change as little as possible throughout the<br />

<strong>in</strong>terval.<br />

4) For each <strong>in</strong>terval i, determ<strong>in</strong>e the neutron fluence Fts1i necessary to accumulate a<br />

l<strong>in</strong>ear swell<strong>in</strong>g stra<strong>in</strong> (1/3 the volumetr<strong>ic</strong> swell<strong>in</strong>g stra<strong>in</strong> <strong>for</strong> isotrop<strong>ic</strong> swell<strong>in</strong>g) of<br />

2% at a temperature of Tmi and a neutron flux of Fmi from A.MAT.4.4 of<br />

appendix A, and the neutron fluence Ftc1i necessary to accumulate an effective <strong>in</strong>reactor<br />

creep stra<strong>in</strong> of 2% at an effective stress of Smi (Tmi,Ftmi ), a temperature<br />

of Tmi and a neutron flux of Fmi, from A.MAT.4.4 of appendix A.<br />

5) Compute the follow<strong>in</strong>g sum of the neutron fluence ratios <strong>for</strong> all N-<strong>in</strong>tervals.<br />

S<br />

N<br />

æ tmi<br />

t<br />

= å ç<br />

+<br />

è t t<br />

F F<br />

F<br />

F<br />

i = 1<br />

mi<br />

ci 1 si 1<br />

ö<br />

÷<br />

ø<br />

6) If the above sum, S, is greater than 1, then a nonl<strong>in</strong>ear analysis <strong>in</strong>volv<strong>in</strong>g large<br />

displacement and stra<strong>in</strong> may be needed.<br />

Note 1: If the material displays an <strong>in</strong>cubation period <strong>for</strong> swell<strong>in</strong>g and the peak neutron<br />

fluence <strong>in</strong> the component is less than the lowest <strong>in</strong>cubation neutron fluence with<strong>in</strong> the range<br />

of temperature (not necessarily the maximum temperature) experienced by the component,<br />

then check<strong>in</strong>g <strong>for</strong> the swell<strong>in</strong>g stra<strong>in</strong> is not needed.<br />

Note 2: Denote the maximum values of the th<strong>ic</strong>kness-averaged temperature, neutron fluence,<br />

and Sm over the whole operat<strong>in</strong>g period (tmax) of the component by Tm, max, Ftm, max and Sm,<br />

max, respectively. If the total creep stra<strong>in</strong> accumulated at a stress Sm, max, temperature Tm,max,<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 21


ITER G 74 MA 8 01-05-28 W0.2<br />

neutron fluence Ftm, max, and average flux Fave (=Ftm, max/tmax) is less than 1%, then<br />

check<strong>in</strong>g of the creep stra<strong>in</strong> is not needed. If there is a peak <strong>in</strong> the irradiation-<strong>in</strong>duced creep<br />

stra<strong>in</strong> at a temperature


ITER G 74 MA 8 01-05-28 W0.2<br />

Note 2: Tables <strong>for</strong> fts are provided <strong>in</strong> A.MAT.4.2 of appendix A.<br />

Note 3: For ITER operat<strong>in</strong>g conditions, this test is satisfied, i.e., S < 1, <strong>for</strong> type 316L(N)<br />

sta<strong>in</strong>less steel and need not be checked.<br />

B 3023 Elast<strong>ic</strong> Analysis<br />

This type of analysis is made assum<strong>in</strong>g that:<br />

- the behaviour of the material is l<strong>in</strong>ear-elast<strong>ic</strong>,<br />

- the material is homogeneous and isotrop<strong>ic</strong>,<br />

- the displacements and stra<strong>in</strong>s are small,<br />

- the <strong>in</strong>itial or residual stresses are zero.<br />

The behaviour of the material is determ<strong>in</strong>ed by the Young's modulus, E, and Poisson's ration,<br />

n, the shear modulus, G, is equal to E/2 (1 + n). The values of Young's modulus as a function<br />

of temperature are given <strong>in</strong> A.MAT.2.2 of appendix A. The value of Poisson's ratio, given <strong>in</strong><br />

A.MAT.2.3 of appendix A, is generally equal to 0.3 but can vary <strong>in</strong> certa<strong>in</strong> cases.<br />

The effects of irradiation-<strong>in</strong>duced swell<strong>in</strong>g may be neglected (B 3022) if the total estimated<br />

maximum irradiation-<strong>in</strong>duced volumetr<strong>ic</strong> swell<strong>in</strong>g stra<strong>in</strong> at the end of <strong>design</strong> life is


ITER G 74 MA 8 01-05-28 W0.2<br />

from all sources should be limited to satisfy the de<strong>for</strong>mation limits <strong>for</strong> functional adequacy<br />

(IC 3040).<br />

S<strong>in</strong>ce irradiation-<strong>in</strong>duced swell<strong>in</strong>g and creep are time-dependent phenomena, de<strong>for</strong>mations<br />

and stress <strong>in</strong>tensities are necessarily time-dependent. Generally, irradiation-<strong>in</strong>duced creep<br />

changes the <strong>in</strong>itial elast<strong>ic</strong> stress distribution by relax<strong>in</strong>g secondary (e.g., thermal or<br />

constra<strong>in</strong>ed swell<strong>in</strong>g stress) or peak stresses with accumulat<strong>in</strong>g neutron fluence. The relaxed<br />

part of the secondary and peak stresses will reappear with reversed sign when the reactor is<br />

shut down and the <strong>in</strong>cremental load<strong>in</strong>g is applied <strong>in</strong> reverse. These changes <strong>in</strong> stress<br />

distribution should be accounted <strong>for</strong> <strong>in</strong> satisfy<strong>in</strong>g fatigue, ratchet, and buckl<strong>in</strong>g limits.<br />

Constitutive equations <strong>for</strong> irradiation-<strong>in</strong>duced creep and swell<strong>in</strong>g <strong>for</strong> the ITER <strong>structural</strong><br />

materials are given <strong>in</strong> the ITER MPH-IV 2 . These equations should be used <strong>in</strong> a f<strong>in</strong>iteelement<br />

elast<strong>ic</strong>-creep analysis of the component. In some cases, simplified constitutive<br />

equations <strong>for</strong> irradiation-<strong>in</strong>duced swell<strong>in</strong>g and creep may be used, provided they can be<br />

shown to give conservative results. Two such cases where stresses evolve with time are<br />

discussed - one <strong>for</strong> swell<strong>in</strong>g-<strong>in</strong>duced stress (B 3024.1.1.1) and the other <strong>for</strong> relaxation of<br />

thermal stress by irradiation-<strong>in</strong>duced creep (B 3024.1.1.2).<br />

B 3024.1.1.1 Swell<strong>in</strong>g <strong>in</strong>duced stress value<br />

If the negligible swell<strong>in</strong>g test of B 3022 is not satisfied, then stresses due to constra<strong>in</strong>ed<br />

irradiation-<strong>in</strong>duced swell<strong>in</strong>g (B 2513) have to be considered. For comput<strong>in</strong>g stresses due to<br />

fluence-dependent swell<strong>in</strong>g, the relax<strong>in</strong>g effects of irradiation-<strong>in</strong>duced creep has to be taken<br />

<strong>in</strong>to account irrespective of whether the negligible irradiation-<strong>in</strong>duced creep test of B 3101 is<br />

satisfied or not, because otherwise the elast<strong>ic</strong>ally calculated swell<strong>in</strong>g stresses could become<br />

unrealist<strong>ic</strong>ally large. Swell<strong>in</strong>g stra<strong>in</strong>s, when constra<strong>in</strong>ed, give rise to stresses that, <strong>in</strong> general,<br />

<strong>in</strong>crease with fluence and ultimately reach steady-state values that are the results of a<br />

dynam<strong>ic</strong> equilibrium between the elast<strong>ic</strong>ally driven constra<strong>in</strong>ed swell<strong>in</strong>g stresses and the<br />

relaxation effects of irradiation-<strong>in</strong>duced creep. Such constra<strong>in</strong>ed swell<strong>in</strong>g stresses are<br />

classified as secondary stresses (ICÊ2525). In general, solv<strong>in</strong>g <strong>for</strong> constra<strong>in</strong>ed swell<strong>in</strong>g<br />

stresses would require an <strong>in</strong>cremental thermal-elast<strong>ic</strong>-creep analysis of the component. For<br />

some isotrop<strong>ic</strong> materials, such as type 316 austenit<strong>ic</strong> sta<strong>in</strong>less steel, the constitutive equations<br />

<strong>for</strong> the fluence driven creep and swell<strong>in</strong>g stra<strong>in</strong>s can be approximated by<br />

and<br />

e<br />

ij creep<br />

( ft)<br />

= BT ( , f)<br />

S<br />

3<br />

2<br />

eij<br />

swell<strong>in</strong>g 1 1 V 1<br />

= d = AT ( , f) d<br />

( ft)<br />

3 V ( ft) 3<br />

where<br />

V = the volume,<br />

ij<br />

ij ij<br />

A(T, f) and B(T, f) are coeff<strong>ic</strong>ients <strong>for</strong> irradiation-<strong>in</strong>duced swell<strong>in</strong>g and<br />

creep equations (see A.MAT.4.2 and A.MAT.4.3 of appendix<br />

A), with T as temperature and f as neutron flux,<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 24<br />

(1)<br />

(2)


ITER G 74 MA 8 01-05-28 W0.2<br />

Sij = the deviator<strong>ic</strong> stress, and<br />

dij = the Kronecker delta.<br />

For these materials, the swell<strong>in</strong>g stresses may be obta<strong>in</strong>ed by conduct<strong>in</strong>g a visco-elast<strong>ic</strong><br />

analysis us<strong>in</strong>g the above equations. Alternatively, the follow<strong>in</strong>g fully constra<strong>in</strong>ed steadystate<br />

swell<strong>in</strong>g stresses may be used as an upper bound.<br />

For the case of uniaxial constra<strong>in</strong>t,<br />

Ds<br />

s<br />

( )<br />

( )<br />

æ AT,<br />

f ö<br />

= - ç<br />

è 3BT,<br />

f<br />

÷<br />

ø .<br />

For the case of biaxial constra<strong>in</strong>t,<br />

Ds<br />

æ 21 ( n) AT,<br />

f ö<br />

= - ç<br />

d a b<br />

è 3BT<br />

f<br />

÷ <strong>for</strong> , = 1,2<br />

, ø<br />

- ( )<br />

( )<br />

abs ab<br />

To reduce the degree of conservatism (e.g., <strong>in</strong> the case of swell<strong>in</strong>g stra<strong>in</strong> gradient), an elast<strong>ic</strong><br />

analysis of the component may be conducted us<strong>in</strong>g an imposed stra<strong>in</strong> distribution def<strong>in</strong>ed as<br />

follows.<br />

For uniaxial load<strong>in</strong>g,<br />

De<br />

s<br />

=<br />

( )<br />

( )<br />

æ AT,<br />

f ö<br />

ç<br />

è 3EB<br />

T,<br />

f<br />

÷<br />

ø<br />

For biaxial load<strong>in</strong>g,<br />

De<br />

=<br />

æ 21 ( - u) 2<br />

AT ( , f)<br />

ö<br />

ç<br />

d a b<br />

è<br />

3EB<br />

T f ÷ <strong>for</strong> , = 1,2<br />

( , )<br />

ø<br />

abs ab<br />

In above, E and n are the Young's modulus and Poisson's ratio at temperature and fluence.<br />

Note that if the mechan<strong>ic</strong>al and phys<strong>ic</strong>al properties are not functions of temperature, the<br />

above stra<strong>in</strong>s may be imposed by a f<strong>ic</strong>titious temperature distribution obta<strong>in</strong>ed by divid<strong>in</strong>g<br />

the stra<strong>in</strong> by the thermal expansion coeff<strong>ic</strong>ient. If the mechan<strong>ic</strong>al and phys<strong>ic</strong>al properties are<br />

functions of temperature, then a method must be found (e.g., a user subrout<strong>in</strong>e) to specify the<br />

swell<strong>in</strong>g stra<strong>in</strong> directly. These secondary stresses together with others contribute to the total<br />

stress (IC 2512). The stress tensor giv<strong>in</strong>g the maximum swell<strong>in</strong>g stress <strong>in</strong>tensity value<br />

among the operat<strong>in</strong>g conditions should be selected.<br />

B 3024.1.1.2 Relaxation of thermal stress by irradiation<strong>in</strong>duced<br />

creep<br />

At low temperatures, where thermal creep is negligible (B 3050), thermal stresses reach a<br />

peak at the beg<strong>in</strong>n<strong>in</strong>g of life and then decrease due to relaxation by irradiation-<strong>in</strong>duced creep.<br />

The analysis of relaxation of thermal stresses by irradiation-<strong>in</strong>duced creep is straight<strong>for</strong>ward,<br />

s<strong>in</strong>ce these stresses are fully developed <strong>in</strong> a very short period of time and then relaxation<br />

takes over. The relaxed part of the thermal stresses will reappear dur<strong>in</strong>g shutdown. Dur<strong>in</strong>g<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 25


ITER G 74 MA 8 01-05-28 W0.2<br />

relaxation, the sum of the <strong>in</strong>cremental elast<strong>ic</strong> stra<strong>in</strong> and <strong>in</strong>cremental irradiation-<strong>in</strong>duced creep<br />

stra<strong>in</strong> is zero, i.e.,<br />

e<br />

ij<br />

c<br />

ij<br />

eÇ + eÇ<br />

= 0 (1)<br />

If the material obeys the creep law given by Eq. 1 of B 3024.1.1.1, then denot<strong>in</strong>g the <strong>in</strong>itial<br />

and relaxed thermal stresses by so and s respectively, Eq. (1) can be solved as follows:<br />

<strong>for</strong> the uniaxial case,<br />

( )<br />

s = soexp -EB(<br />

T, f) ft<br />

(2)<br />

and <strong>for</strong> the equi-biaxial case,<br />

æ 3 EB( T, f) ftö<br />

sij = sodij expç<br />

-<br />

÷<br />

è 2 ( 1 - n)<br />

ø<br />

where i , j = 1,2<br />

and E and n are the Young's modulus and Poisson's ratio.<br />

If the material obeys a creep law different from Eq. (1) of B 3024.1.1.1 or if the <strong>in</strong>itial<br />

start<strong>in</strong>g stress is more general than equi-biaxial, Eq(1) can be solved numer<strong>ic</strong>ally if a closed<br />

<strong>for</strong>m solution cannot be obta<strong>in</strong>ed.<br />

B 3024.1.2 Neuber's rule<br />

Neuber's rule can be applied to estimate the maximum elasto-plast<strong>ic</strong> stresses and stra<strong>in</strong>s at<br />

notch roots. Consider a notch with an elast<strong>ic</strong> stress concentration factor KT subjected to a<br />

nom<strong>in</strong>al (remote) uniaxial stress So and, <strong>for</strong> a l<strong>in</strong>ear elast<strong>ic</strong> material, correspond<strong>in</strong>g remote<br />

uniaxial stra<strong>in</strong> eo (eo = So/E). The elast<strong>ic</strong>ally calculated peak stress (S) and stra<strong>in</strong> (e) at the<br />

notch root are given by<br />

S = KTSo<br />

e = KTeo<br />

NeuberÕs rule states that, if we replace the l<strong>in</strong>ear elast<strong>ic</strong> material with a material obey<strong>in</strong>g a<br />

uniaxial power-law constitutive equation,<br />

s = Ae n ,<br />

then, denot<strong>in</strong>g the notch root maximum stress and stra<strong>in</strong> by s and e,<br />

s · e = S · e (1)<br />

The above equations can be solved <strong>for</strong> the maximum stress and stra<strong>in</strong> at the notch roots as<br />

s = SK<br />

e = eK<br />

2n/( 1+<br />

n)<br />

o T<br />

2/( 1+<br />

n)<br />

o T<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 26<br />

(3)<br />

(2a)<br />

(2b)


ITER G 74 MA 8 01-05-28 W0.2<br />

Equations 2a-b have also been used extensively <strong>in</strong> fatigue analysis of notches by replac<strong>in</strong>g<br />

the stresses and stra<strong>in</strong>s by their respective ranges and by replac<strong>in</strong>g KT with Keff, wh<strong>ic</strong>h is an<br />

experimentally determ<strong>in</strong>ed (material-dependent) effective stress concentration factor.<br />

Neuber's rule <strong>in</strong> the <strong>for</strong>m of Eq. 1 is used <strong>in</strong> satisfy<strong>in</strong>g the fatigue damage limit (IC 3322)<br />

us<strong>in</strong>g the elast<strong>ic</strong> fatigue analysis rule (B 3323.1).<br />

Note: Neuber's rule as described here applies str<strong>ic</strong>tly to a power-law harden<strong>in</strong>g material,<br />

wh<strong>ic</strong>h implies that, <strong>in</strong> order to be appl<strong>ic</strong>able to an elast<strong>ic</strong>-power-law harden<strong>in</strong>g material, the<br />

elast<strong>ic</strong> stra<strong>in</strong> at the notch root must be small compared to the plast<strong>ic</strong> stra<strong>in</strong>. Also, <strong>in</strong> a<br />

multiaxial load<strong>in</strong>g situation, the stresses and stra<strong>in</strong>s have to be replaced by their respective<br />

scalar representations.<br />

B 3024.1.3 Elast<strong>ic</strong> follow-up factor (r)<br />

The actual stress and stra<strong>in</strong> at any po<strong>in</strong>t <strong>in</strong> an elasto-plast<strong>ic</strong>ally de<strong>for</strong>m<strong>in</strong>g structure can be<br />

expressed <strong>in</strong> terms of the elast<strong>ic</strong>ally calculated stress at the same po<strong>in</strong>t by the use of a factor<br />

R, def<strong>in</strong>ed as follows:<br />

Ee<br />

- s<br />

R =<br />

s - s<br />

where<br />

el<br />

E = Young's modulus,<br />

sel = elast<strong>ic</strong>ally calculated stress<br />

s and e = actual stress and total stra<strong>in</strong><br />

The maximum value of R <strong>in</strong> the structure, wh<strong>ic</strong>h occurs at the po<strong>in</strong>t of maximum stress or<br />

plast<strong>ic</strong> stra<strong>in</strong>, is called the Òelast<strong>ic</strong> follow-up factorÓ, r. The term elast<strong>ic</strong> follow-up (IC<br />

2161) refers to the fact that the total stra<strong>in</strong> <strong>in</strong>cludes plast<strong>ic</strong> stra<strong>in</strong>s, wh<strong>ic</strong>h tend to be driven or<br />

ÒfollowedÓ by the stra<strong>in</strong> energy <strong>in</strong> the elast<strong>ic</strong> part of the structure. In the l<strong>iter</strong>ature (e.g.,<br />

Roche), the elast<strong>ic</strong> follow up factor has sometimes been def<strong>in</strong>ed <strong>in</strong> terms of a different factor<br />

rÕ wh<strong>ic</strong>h is numer<strong>ic</strong>ally equal to r Ð 1, with r def<strong>in</strong>ed as above. In general, the value of r<br />

depends on the geometry of the structure, material stress-stra<strong>in</strong> law, and the load level. When<br />

def<strong>in</strong>ed as <strong>in</strong> Eq. (1), the maximum plast<strong>ic</strong> stra<strong>in</strong> <strong>in</strong> the structure (ep) can be expressed as<br />

r<br />

ep= sel - s<br />

E<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 27<br />

(1)<br />

[ ] (2)<br />

Quite often, the actual stress s is much less than either Ee or the elast<strong>ic</strong>ally calculated stress<br />

sel, <strong>in</strong> wh<strong>ic</strong>h case the r-factor can be shown to be equal to the stra<strong>in</strong> concentration factor due<br />

to elast<strong>ic</strong> follow-up, i.e.,<br />

r<br />

E e e<br />

= = =<br />

s<br />

e<br />

el el<br />

K<br />

e<br />

A signif<strong>ic</strong>ant number of detailed <strong>in</strong>elast<strong>ic</strong> analyses have been reported <strong>in</strong> the l<strong>iter</strong>ature<br />

<strong>in</strong>d<strong>ic</strong>at<strong>in</strong>g that, <strong>for</strong> general structures made of a reasonably stra<strong>in</strong>-harden<strong>in</strong>g ductile material<br />

(3)


ITER G 74 MA 8 01-05-28 W0.2<br />

and with maximum nom<strong>in</strong>al stresses (P+Q) not exceed<strong>in</strong>g the 3Sm limit (IC 3311.1), a<br />

conservative value of r = 3. The <strong>design</strong> guide <strong>for</strong> the Monju reactor of Japan recommends a<br />

default value of r=3 <strong>for</strong> creep-fatigue <strong>design</strong>. However, the r-factor methodology has not<br />

been used <strong>in</strong> the fission reactor codes to <strong>design</strong> aga<strong>in</strong>st fracture, because loss of ductility is<br />

not a damage considered <strong>in</strong> these codes. S<strong>in</strong>ce, <strong>in</strong> the SDC-IC, we propose to use the r-factor<br />

methodology to <strong>design</strong> aga<strong>in</strong>st fracture, two factors have to be considered, both of wh<strong>ic</strong>h tend<br />

to <strong>in</strong>crease the value of r. They are (1) reduced or zero stra<strong>in</strong> harden<strong>in</strong>g capability of<br />

materials under neutron irradiation (see Figure C 3024-6 of Appendix C) and (2) the elast<strong>ic</strong><br />

stress that corresponds to fracture far exceeds the 3Sm limit.<br />

Analysis of displacement-controlled three-po<strong>in</strong>t bend tests (see C 3024.1.3 of Appendix C) of<br />

a beam made of a material with a bil<strong>in</strong>ear stress-stra<strong>in</strong> law has shown that the value of r as a<br />

function of the peak plast<strong>ic</strong> stra<strong>in</strong> is bounded (£ 4) as long as the ratio of the tangent modulus<br />

to Young's modulus (ET/E) is ³ 0.01. At lower values of ET/E ratio, the r value <strong>in</strong>creases<br />

with <strong>in</strong>creas<strong>in</strong>g peak plast<strong>ic</strong> stra<strong>in</strong> and, <strong>in</strong> part<strong>ic</strong>ular, <strong>for</strong> an elast<strong>ic</strong>-perfectly plast<strong>ic</strong> material<br />

(ET/E = 0), it <strong>in</strong>creases <strong>in</strong>def<strong>in</strong>itely with <strong>in</strong>creas<strong>in</strong>g peak plast<strong>ic</strong> stra<strong>in</strong>.<br />

Experimental results on three-po<strong>in</strong>t bend tests on irradiated type 304 sta<strong>in</strong>less steel have<br />

shown that the value of r can be large if the uni<strong>for</strong>m elongation of the material is £ 1% (see C<br />

3024.1.4 of Appendix C). When extrapolated from this set of data, the value of r should be £<br />

4 when the uni<strong>for</strong>m elongation ³ 2%. These results are <strong>in</strong> good agreement with the<br />

analyt<strong>ic</strong>ally determ<strong>in</strong>ed r as discussed <strong>in</strong> the previous paragraph.<br />

In contrast to the three-po<strong>in</strong>t bend tests, s<strong>in</strong>gle-edge notched tensile tests on irradiated type<br />

304 sta<strong>in</strong>less steel have shown that the value of r is relatively constant (» KT) even when the<br />

uni<strong>for</strong>m elongation is < 1% (see C 3024.1.3 of Appendix C). These results are <strong>in</strong> agreement<br />

with r-values pred<strong>ic</strong>ted analyt<strong>ic</strong>ally on the assumption that stra<strong>in</strong> localization at the notch<br />

cannot occur and that plast<strong>ic</strong> stra<strong>in</strong>s rema<strong>in</strong> distributed homogeneously at the notch root (see<br />

C 3024.1.4 of Appendix C). However, because of the stiffness of the test<strong>in</strong>g mach<strong>in</strong>e, these<br />

specimens did not have the freedom to fail by slid<strong>in</strong>g off at an angle start<strong>in</strong>g at the notch. In<br />

a load controlled test, depend<strong>in</strong>g on the notch geometry, stra<strong>in</strong> localization could occur. This<br />

must either be prevented by restr<strong>ic</strong>t<strong>in</strong>g the range of uni<strong>for</strong>m elongation or accounted <strong>for</strong> by<br />

<strong>in</strong>creas<strong>in</strong>g the r factor.<br />

It is clear that the value of r can be quite sensitive to the component geometry and load<strong>in</strong>g<br />

mode. There<strong>for</strong>e, <strong>in</strong> the <strong>design</strong> rules (IC 3212.1 and 3213.1), wh<strong>ic</strong>h use the r-factor<br />

methodology <strong>for</strong> sett<strong>in</strong>g the limits on elast<strong>ic</strong>ally calculated maximum stresses, the value of r<br />

has been chosen conservatively. However, <strong>in</strong> any part<strong>ic</strong>ular appl<strong>ic</strong>ation, the <strong>design</strong>er is given<br />

the option of us<strong>in</strong>g a lower value of r if it can be so justified.<br />

The justif<strong>ic</strong>ation <strong>for</strong> a different r consists of estimat<strong>in</strong>g the peak values of stra<strong>in</strong> and stress<br />

conservatively us<strong>in</strong>g elasto-plast<strong>ic</strong> analysis (B 3024.2), us<strong>in</strong>g Eq. 1 to evaluate R, and tak<strong>in</strong>g<br />

r as the maximum R <strong>in</strong> the structure. The <strong>design</strong>er should use judgment <strong>in</strong> select<strong>in</strong>g a<br />

suitable sub-model and load<strong>in</strong>g of the structure. See BÊ3025 <strong>for</strong> a discussion of calculation<br />

zones and sub-models. In addition, the <strong>design</strong>er should use an appropriate stress-stra<strong>in</strong> law<br />

that corresponds to the lowest work harden<strong>in</strong>g (wh<strong>ic</strong>h, <strong>for</strong> steel, occurs at the maximum<br />

fluence) under consideration. The analysis must be carried out either to the load level of<br />

<strong>in</strong>terest or to demonstrate that the r-value has peaked prior to reach<strong>in</strong>g the load level of<br />

<strong>in</strong>terest.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 28


ITER G 74 MA 8 01-05-28 W0.2<br />

BÊ3024.2 Elasto-plast<strong>ic</strong> analysis of a structure subjected to a<br />

monoton<strong>ic</strong> load<strong>in</strong>g<br />

Elasto-plast<strong>ic</strong> analysis should be conducted by us<strong>in</strong>g a validated f<strong>in</strong>ite-element analysis<br />

program (e.g., ANSYS, ABAQUS, etc.) together with a validated material model<br />

(constitutive law) and a <strong>structural</strong> model of quantified accuracy. Alternatively, the <strong>design</strong>er<br />

may use any other analyt<strong>ic</strong>al or numer<strong>ic</strong>al means that can be justified. Generally, an elastoplast<strong>ic</strong><br />

analysis will be required under two conditions. First, if the <strong>design</strong>er <strong>in</strong>tends to use a<br />

lower value of the elast<strong>ic</strong> follow-up factor r than is stipulated <strong>in</strong> the elast<strong>ic</strong> analysis rules of<br />

IC 3211.1, 3212.1, and 3213.1, the value of r has to be justified on the basis of elasto-plast<strong>ic</strong><br />

analyses (B 3024.2). Second, if the <strong>design</strong>er cannot satisfy the elast<strong>ic</strong> analysis rules of either<br />

IC 3211.1, 3212.1, or 3213.1, the elasto-plast<strong>ic</strong> analysis rules of IC 3211.2, 3212.2 or 3213.2<br />

have to be satisfied us<strong>in</strong>g elasto-plast<strong>ic</strong> analyses.<br />

This section describes a simplified elasto-plast<strong>ic</strong> analysis procedure appl<strong>ic</strong>able to monoton<strong>ic</strong><br />

load<strong>in</strong>g. This is appl<strong>ic</strong>able only if it can be ascerta<strong>in</strong>ed that the follow<strong>in</strong>g two conditions are<br />

satisfied:<br />

1) no unload<strong>in</strong>g ever occurs <strong>in</strong> any part of the structure wh<strong>ic</strong>h undergoes plast<strong>ic</strong><br />

stra<strong>in</strong>,<br />

2) at all po<strong>in</strong>ts of the structure, the stress tensor <strong>components</strong> calculated elast<strong>ic</strong>ally<br />

vary simultaneously and proportionally to a s<strong>in</strong>gle parameter.<br />

A monoton<strong>ic</strong> load<strong>in</strong>g is one wh<strong>ic</strong>h evolves <strong>in</strong> a constantly <strong>in</strong>creas<strong>in</strong>g or decreas<strong>in</strong>g manner.<br />

For this type of analysis, the mathemat<strong>ic</strong>al model of the material behaviour is based on the<br />

follow<strong>in</strong>g laws:<br />

- homogeneity and isotrop<strong>ic</strong> behaviour of the material,<br />

- von Mises yield cr<strong>iter</strong>ion,<br />

- the associated normality law <strong>for</strong> plast<strong>ic</strong> stra<strong>in</strong>: plast<strong>ic</strong> flow rule,<br />

- an isotrop<strong>ic</strong> stra<strong>in</strong> harden<strong>in</strong>g rule.<br />

The material characterist<strong>ic</strong>s required <strong>for</strong> appl<strong>ic</strong>ation of this model are:<br />

- the m<strong>in</strong>imum tensile curves as a function of temperature and fluence (A1.6.1),<br />

- the Young's modulus and Poisson's ratio as a function of temperature and fluence<br />

(A1.2.2).<br />

The elasto-plast<strong>ic</strong> analysis rules of IC 3211.1, 3212.2, and 3213.2, require the use of various<br />

load factors (GiL) to be applied to mechan<strong>ic</strong>al loads, and stra<strong>in</strong> factors (GiS) to be applied to<br />

de<strong>for</strong>mation-controlled loads (e.g., thermal load, swell<strong>in</strong>g-<strong>in</strong>duced load) be<strong>for</strong>e the elastoplast<strong>ic</strong><br />

analysis is carried out. Although the appl<strong>ic</strong>ation of a load factor to mechan<strong>ic</strong>al loads<br />

is straight<strong>for</strong>ward, <strong>for</strong> thermally-<strong>in</strong>duced (or swell<strong>in</strong>g-<strong>in</strong>duced) load<strong>in</strong>g, the stra<strong>in</strong> factor is<br />

applied to the loads <strong>in</strong>duced by thermal stra<strong>in</strong>s (or swell<strong>in</strong>g stra<strong>in</strong>s). In the latter case, it may<br />

be necessary to artif<strong>ic</strong>ially <strong>in</strong>duce high stra<strong>in</strong>s concurrent with the use of realist<strong>ic</strong> stiffness<br />

and flow properties. The use of an "adjusted" thermal expansion coeff<strong>ic</strong>ient is one technique<br />

<strong>for</strong> enhanc<strong>in</strong>g the applied stra<strong>in</strong>s without affect<strong>in</strong>g the associated stiffness and flow<br />

properties. The treatment of swell<strong>in</strong>g-<strong>in</strong>duced stress is discussed <strong>in</strong> B 3024.1.1.1.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 29


ITER G 74 MA 8 01-05-28 W0.2<br />

BÊ3024.2.1 Use of the tensile curve<br />

To characterize the stress-stra<strong>in</strong> response, this elasto-plast<strong>ic</strong> model makes use of a Òm<strong>in</strong>imum<br />

tensile curve,Ó expressed as a relationship between the stress s* and the cumulative plast<strong>ic</strong><br />

stra<strong>in</strong> e p * at a given temperature. This relationship is obta<strong>in</strong>ed by subtract<strong>in</strong>g the elast<strong>ic</strong><br />

stra<strong>in</strong> from the total stra<strong>in</strong> of the tensile curve as shown <strong>in</strong> Figure BÊ3024-1 (below). The<br />

m<strong>in</strong>imum (lower bound) tensile curve is used to obta<strong>in</strong> a conservatively high estimate of the<br />

plast<strong>ic</strong> stra<strong>in</strong>.<br />

*<br />

The curve ( s*; ep)<br />

represents the stra<strong>in</strong> harden<strong>in</strong>g behaviour of the material. s* depends<br />

*<br />

on the current stra<strong>in</strong> harden<strong>in</strong>g state as represented by the cumulative plast<strong>ic</strong> stra<strong>in</strong> ep . In the<br />

<strong>in</strong>itial virg<strong>in</strong> state of the material, e p * = 0, s* equals the <strong>in</strong>itial yield stress.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 30


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure BÊ3024-1: Use of tensile curve<br />

BÊ3024.2.2 Plast<strong>ic</strong>ity cr<strong>iter</strong>ion<br />

The von Mises yield cr<strong>iter</strong>ion can be expressed <strong>in</strong> terms of the stress <strong>in</strong>tensity def<strong>in</strong>ed <strong>in</strong><br />

BÊ3254.1 and the stress s* determ<strong>in</strong>ed <strong>in</strong> BÊ3024.2.1:<br />

* *<br />

s = s ( ep)<br />

where<br />

s = s s<br />

3<br />

© ij © ij<br />

2<br />

where s'ij is the deviator<strong>ic</strong> stress. Alternatively,<br />

2<br />

2<br />

{ ( ) + ( 2 - 3)<br />

+ ( 3 - 1)<br />

}<br />

s = . s - s s s s s<br />

12 1 2<br />

The behaviour of the material rema<strong>in</strong>s elast<strong>ic</strong> as long as s < s*.<br />

When s = s*,<br />

the limit<br />

of the elast<strong>ic</strong> behaviour is reached and plast<strong>ic</strong> stra<strong>in</strong> can occur.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 31<br />

2 12


ITER G 74 MA 8 01-05-28 W0.2<br />

p<br />

Plast<strong>ic</strong> stra<strong>in</strong> eij is calculated <strong>in</strong>crementally by us<strong>in</strong>g the flow rule given <strong>in</strong> ICÊ3024.2.3. In<br />

the case of a material wh<strong>ic</strong>h can stra<strong>in</strong> harden, the plast<strong>ic</strong> stra<strong>in</strong> modifies the yield cr<strong>iter</strong>ion.<br />

This modif<strong>ic</strong>ation is determ<strong>in</strong>ed by means of the harden<strong>in</strong>g rule given <strong>in</strong> ICÊ3024.2.4.<br />

BÊ3024.2.3 Flow rule<br />

p<br />

The flow rule provides the <strong>in</strong>cremental plast<strong>ic</strong> stra<strong>in</strong> de ij <strong>for</strong> a given <strong>in</strong>cremental stress dsij.<br />

The <strong>in</strong>cremental plast<strong>ic</strong> stra<strong>in</strong> at any po<strong>in</strong>t can be obta<strong>in</strong>ed by us<strong>in</strong>g the follow<strong>in</strong>g steps.<br />

1) At the start of a step, the current stress, stra<strong>in</strong>, and plast<strong>ic</strong> stra<strong>in</strong> at a given po<strong>in</strong>t<br />

is known. By def<strong>in</strong>ition, the current stress is on the current yield surface wh<strong>ic</strong>h<br />

is known:<br />

2<br />

3<br />

* * 2<br />

f( s©, ij s ) s© ij s© ij s<br />

where<br />

= - ( ) =<br />

s'ij is the deviator<strong>ic</strong> stress.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 32<br />

0,<br />

2) For a given stress <strong>in</strong>crement dsij , the deviator<strong>ic</strong> stress <strong>in</strong>crement ds'ij can be<br />

calculated. The po<strong>in</strong>t under consideration will undergo further plast<strong>ic</strong><br />

de<strong>for</strong>mation if s'ij ds'ij > 0. The stress <strong>in</strong>crement will cause an elast<strong>ic</strong> stra<strong>in</strong><br />

<strong>in</strong>crement only if s'ij ds'ij £ 0.<br />

3) If further plast<strong>ic</strong> de<strong>for</strong>mation occurs, the plast<strong>ic</strong> stra<strong>in</strong> <strong>in</strong>crement is determ<strong>in</strong>ed<br />

by the normality rule (Reuss equations)<br />

de<br />

p<br />

ij<br />

f<br />

= dl = s© ij dl<br />

s©<br />

ij<br />

.<br />

Def<strong>in</strong><strong>in</strong>g an equivalent plast<strong>ic</strong> stra<strong>in</strong> <strong>in</strong>crement by<br />

p<br />

de =<br />

p p<br />

deijdeij 2<br />

,<br />

3<br />

wh<strong>ic</strong>h can be written <strong>in</strong> an expanded <strong>for</strong>m as<br />

( ) + ( - ) + ( - )<br />

2<br />

de =<br />

ì<br />

2<br />

2<br />

í de11 - de22 de22 de33 de de<br />

3 î<br />

12<br />

p p p<br />

+<br />

æ 2 2 2<br />

( de ) + ( de ) + ( de<br />

ö ü<br />

6<br />

è 12 23 31)<br />

ø ý<br />

þ<br />

p p p p p p p 2<br />

33 11<br />

dl can be solved as<br />

dl<br />

d p<br />

3 e<br />

= *<br />

2 s ,<br />

and the <strong>in</strong>cremental plast<strong>ic</strong> stra<strong>in</strong> <strong>components</strong> are then given by :<br />

,


ITER G 74 MA 8 01-05-28 W0.2<br />

de = s - 05 . s + s de<br />

s<br />

de = s - 05 . s + s de<br />

s<br />

de = s - 05 . s + s de<br />

s<br />

de 32 s de<br />

s<br />

de 32 s de<br />

s<br />

de<br />

p<br />

11 11 22 33<br />

p<br />

22 22 33 11<br />

p<br />

33 33 11 22<br />

p<br />

12 12<br />

p<br />

23 23<br />

p<br />

31<br />

[ ( ) ]<br />

[ ( ) ]<br />

[ ( ) ]<br />

= ( )<br />

= ( )<br />

= ( 32)<br />

p<br />

p<br />

*<br />

*<br />

s31 de s<br />

p *<br />

4) The value of de p has to be determ<strong>in</strong>ed <strong>iter</strong>atively by requir<strong>in</strong>g that the resultant<br />

<strong>in</strong>cremental harden<strong>in</strong>g ds * (obta<strong>in</strong>ed from the tensile curve) and the new stress<br />

state must be on the new yield surface wh<strong>ic</strong>h is determ<strong>in</strong>ed by the harden<strong>in</strong>g rule<br />

(BÊ3024.2.4) and wh<strong>ic</strong>h, <strong>for</strong> isotrop<strong>ic</strong> harden<strong>in</strong>g, is given by,<br />

* *<br />

f ( s© ij + ds©, ij s + ds<br />

) = 0<br />

Note 1: The total stress tensor must also satisfy the equilibrium equations and the total<br />

stra<strong>in</strong>s must satisfy the compatibility equations.<br />

Note 2: In cases where the stress and stra<strong>in</strong> ratios are ma<strong>in</strong>ta<strong>in</strong>ed constant (radial load<strong>in</strong>g),<br />

the <strong>in</strong>cremental stress-stra<strong>in</strong> relationship (Reuss equations) can be <strong>in</strong>tegrated to give a<br />

relationship between total plast<strong>ic</strong> stra<strong>in</strong> and stress (Hencky's equations). Although this may<br />

simplify the calculations signif<strong>ic</strong>antly, it must be remembered that Hencky's equations are<br />

<strong>in</strong>correct <strong>for</strong> general nonradial load<strong>in</strong>g.<br />

BÊ3024.2.4 Harden<strong>in</strong>g rule<br />

The harden<strong>in</strong>g rule used dur<strong>in</strong>g analysis of a structure subjected to monoton<strong>ic</strong> load<strong>in</strong>g is an<br />

isotrop<strong>ic</strong> harden<strong>in</strong>g rule. Accord<strong>in</strong>g to this rule, the yield surface expands but reta<strong>in</strong>s its<br />

shape and orig<strong>in</strong>. The expansion of the yield surface is determ<strong>in</strong>ed by the variation of s*.<br />

After calculat<strong>in</strong>g d p<br />

e by an <strong>iter</strong>ative procedure and determ<strong>in</strong><strong>in</strong>g the new value of the<br />

*<br />

cumulative plast<strong>ic</strong> stra<strong>in</strong> ep , a new value of s* is obta<strong>in</strong>ed with the curve <strong>in</strong> Figure BÊ3024-<br />

1. It should be noted that s* is always positive <strong>in</strong> this model.<br />

For monoton<strong>ic</strong> proportional load<strong>in</strong>g, k<strong>in</strong>emat<strong>ic</strong> harden<strong>in</strong>g gives the same results.<br />

BÊ3024.3 Elasto-plast<strong>ic</strong> analysis of a structure subjected to cycl<strong>ic</strong><br />

load<strong>in</strong>g<br />

A large number of possible analysis methods are now available, but none of them are today<br />

qualified to describe all aspects of the cycl<strong>ic</strong> behaviour of materials. The imposition of any<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 33<br />

p<br />

p<br />

p<br />

*<br />

*<br />

*


ITER G 74 MA 8 01-05-28 W0.2<br />

one of them would not appear necessary. Nevertheless, the used method used must be<br />

justified.<br />

B 3024.4 Elasto-visco-plast<strong>ic</strong> analysis of a structure subjected to cycl<strong>ic</strong><br />

load<strong>in</strong>g<br />

A large number of possible analysis methods are now available, but none of them are today<br />

qualified to describe all aspects of the cycl<strong>ic</strong> behavior of materials. The imposition of any<br />

one of them would not appear necessary. Nevertheless, the method used must be justified.<br />

Note: A comb<strong>in</strong>ed elasto-visco-plast<strong>ic</strong> analysis is necessary only when the effects of creep<br />

and plast<strong>ic</strong>ity cannot be separated from one another. In many cases it is possible to separate<br />

the effects as follows. Most models of plast<strong>ic</strong>ity are <strong>in</strong>dependent of time. If the time frame<br />

<strong>for</strong> plast<strong>ic</strong> sta<strong>in</strong>s is rapid compared with the time frame <strong>for</strong> creep, then one can assume that<br />

creep stra<strong>in</strong>s are constant dur<strong>in</strong>g a plast<strong>ic</strong> load<strong>in</strong>g period and that plast<strong>ic</strong> stra<strong>in</strong>s are constant<br />

dur<strong>in</strong>g the creep periods between plast<strong>ic</strong> load<strong>in</strong>gs. If plast<strong>ic</strong>ity and creep effects are not<br />

separable, then the <strong>for</strong>mulation of the constitutive law becomes more diff<strong>ic</strong>ult.<br />

BÊ3024.5 Limit analysis (collapse load)<br />

The de<strong>for</strong>mation of a structure made of a rigid-perfectly plast<strong>ic</strong> material <strong>in</strong>creases without<br />

bound at a load<strong>in</strong>g level called the collapse load. Limit analysis methods can be used to<br />

calculate the collapse load or a lower bound to the collapse load.<br />

A given load<strong>in</strong>g is less than or equal to the collapse load if there is a stress distribution wh<strong>ic</strong>h<br />

satisfies the laws of equilibrium at all po<strong>in</strong>ts that does not violate the material yield cr<strong>iter</strong>ion<br />

at any po<strong>in</strong>t. This theorem allows a lower bound to be def<strong>in</strong>ed <strong>for</strong> the collapse load.<br />

In the case of elasto-plast<strong>ic</strong> analysis and experimental analysis, the collapse load, by<br />

convention, is def<strong>in</strong>ed as the load<strong>in</strong>g <strong>for</strong> wh<strong>ic</strong>h the overall permanent de<strong>for</strong>mation of the<br />

structure equals the de<strong>for</strong>mation wh<strong>ic</strong>h would occur by purely elast<strong>ic</strong> behavior.<br />

BÊ3025 Zones of calculation<br />

In some cases it may be necessary <strong>for</strong> computational purposes to divide a component <strong>in</strong>to<br />

several zones <strong>for</strong> analyz<strong>in</strong>g a s<strong>in</strong>gle type of damage. In such cases, an overall analysis of the<br />

component shall be carried out to determ<strong>in</strong>e the loads or displacements to be applied at the<br />

boundaries of each zone <strong>for</strong> each load case considered. Care should be taken to ensure that<br />

boundary conditions applied are suff<strong>ic</strong>iently accurate or conservative. The accuracy may be<br />

checked by a comb<strong>in</strong>ation of the follow<strong>in</strong>g:<br />

a. show<strong>in</strong>g that both stresses and displacements are consistent at the boundaries of<br />

the zone;<br />

b. ref<strong>in</strong>ement of the mesh.<br />

Note: the appl<strong>ic</strong>ation of displacement boundary conditions to a sub-model requires care.<br />

Two examples are given below.<br />

Say that an elast<strong>ic</strong> analysis of a larger model is used to determ<strong>in</strong>e the displacement boundary<br />

conditions <strong>for</strong> a smaller sub-model, yet the sub-model is expected to operate <strong>in</strong> the plast<strong>ic</strong><br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 34


ITER G 74 MA 8 01-05-28 W0.2<br />

range. In this case, the elast<strong>ic</strong> approximation of the sub-model region used to def<strong>in</strong>e the<br />

boundary conditions will be too stiff, so the calculated boundary displacements will be too<br />

small, and the plast<strong>ic</strong> stra<strong>in</strong>s <strong>in</strong> the sub-model will be underestimated. The problem will be<br />

much less severe if the plast<strong>ic</strong> zone <strong>in</strong> the sub-model is completely embedded <strong>in</strong>side the<br />

elast<strong>ic</strong> material.<br />

Similar problems can occur with mesh ref<strong>in</strong>ement. If a coarse mesh model is used to def<strong>in</strong>e<br />

displacement boundary conditions <strong>for</strong> a f<strong>in</strong>e mesh sub-model, the coarse mesh will be too<br />

stiff, and stresses <strong>in</strong> the f<strong>in</strong>e mesh sub-model will be underestimated.<br />

BÊ3026 Comb<strong>in</strong>ation of analysis methods<br />

As a general rule, <strong>for</strong> reasons discussed above, the same analysis method (elast<strong>ic</strong> or <strong>in</strong>elast<strong>ic</strong>)<br />

must be used <strong>for</strong> all parts of a component. A method other than that used <strong>for</strong> the entire<br />

apparatus may be used locally provided the results of the two analysis methods are shown to<br />

be consistent (with regard to both stresses and displacements) at the boundaries of the parts<br />

exam<strong>in</strong>ed.<br />

B3030 Appl<strong>ic</strong>able rules - Flow of analysis<br />

This section provides guidel<strong>in</strong>es <strong>for</strong> on the procedure <strong>for</strong> conduct<strong>in</strong>g analyses required <strong>for</strong><br />

satisfy<strong>in</strong>g the <strong>design</strong> rules of SDC-IC.<br />

B3031 Master flow charts <strong>for</strong> satisfy<strong>in</strong>g <strong>design</strong> rules<br />

The flow chart <strong>for</strong> satisfy<strong>in</strong>g the <strong>design</strong> rules of SDC-IC is given <strong>in</strong> Figure IC 3030-1. Three<br />

different event classes are considered <strong>in</strong> the flow chart. Def<strong>in</strong>itions of these event classes are<br />

given <strong>in</strong> IC 2220.<br />

The flow chart <strong>for</strong> satisfy<strong>in</strong>g the low temperature <strong>design</strong> rules <strong>for</strong> a given operat<strong>in</strong>g<br />

conditions is given <strong>in</strong> Figure IC3030-2. In order to use this flow chart, the negligible thermal<br />

creep test of IC 3050 must first be satisfied. An additional flow chart <strong>for</strong> satisfy<strong>in</strong>g high<br />

temperature <strong>design</strong> rules, when the test IC 3050 is not satisfied, will be provided <strong>in</strong> the future.<br />

A more detailed flow chart <strong>for</strong> satisfy<strong>in</strong>g the limits <strong>for</strong> a given Cr<strong>iter</strong>ia Level and <strong>for</strong> a given<br />

damage mode, us<strong>in</strong>g various analysis options, is given <strong>in</strong> Figure B 3030-1.<br />

In general, the degree of adherence to the flowchart procedures and the level of sophist<strong>ic</strong>ation<br />

of analysis are expected to vary with the <strong>design</strong> phase, be<strong>in</strong>g lesser <strong>in</strong> the conceptual <strong>design</strong><br />

phase and greater <strong>in</strong> the f<strong>in</strong>al <strong>design</strong> phase.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 35


ITER G 74 MA 8 01-05-28 W0.2<br />

Start<br />

Conduct Elast<strong>ic</strong>-<br />

Irradiation Creep<br />

Analysis<br />

(B 3024.1.1)<br />

no<br />

Select Damage Mode<br />

Select Cr<strong>iter</strong>ia Level<br />

Check if f<strong>in</strong>ite<br />

de<strong>for</strong>mation<br />

analysis necessary<br />

(B 3021)<br />

Is swell<strong>in</strong>g negligible?<br />

(B 3022)<br />

yes<br />

Is Irradiation-Creep<br />

negligible?<br />

(B 3101)<br />

Conduct Elast<strong>ic</strong><br />

Stress Analysis<br />

(B 3023)<br />

Elast<strong>ic</strong> Analysis<br />

Limits satisfied?<br />

Cr<strong>iter</strong>ia Satisfied<br />

Stop<br />

yes<br />

yes<br />

no<br />

no<br />

Include swell<strong>in</strong>g stress<br />

<strong>in</strong> analysis<br />

(B 3024.1.1.1)<br />

optional<br />

Is Irradiation-Creep<br />

negligible?<br />

(B 3101)<br />

Conduct Elasto-<br />

Plast<strong>ic</strong> Analysis<br />

(B 3024.2)<br />

Elasto-Plast<strong>ic</strong><br />

Analysis Limits<br />

Satisfied?<br />

no<br />

Cr<strong>iter</strong>ia not satisfied<br />

Re<strong>design</strong><br />

Conduct Elasto-<br />

Visco-Plast<strong>ic</strong><br />

Analysis<br />

(B 3024.5)<br />

Figure BÊ3030-1 Analysis flow chart <strong>for</strong> satisfy<strong>in</strong>g the limits of a given<br />

Cr<strong>iter</strong>ia Level and damage mode at low temperatures<br />

B 3040 Rules <strong>for</strong> the prevention of excessive de<strong>for</strong>mation affect<strong>in</strong>g<br />

functional adequacy<br />

yes<br />

BÊ3050 Negligible thermal creep test<br />

For a component or a part of a component, thermal creep is negligible over the total operat<strong>in</strong>g<br />

period if the condition of the test per<strong>for</strong>med us<strong>in</strong>g the follow<strong>in</strong>g procedure is satisfied.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 36<br />

yes<br />

no


ITER G 74 MA 8 01-05-28 W0.2<br />

1) The total operat<strong>in</strong>g period of the component, throughout the life of the component<br />

is taken <strong>in</strong>to account.<br />

2) The total operat<strong>in</strong>g period is divided <strong>in</strong>to N time <strong>in</strong>tervals. For each <strong>in</strong>terval i, of a<br />

duration ti, the maximum temperature reached is denoted Ti,<br />

3) To take full advantage of this rule, the <strong>in</strong>tervals of time should be chosen <strong>in</strong> such a<br />

way that the temperature changes as little as possible throughout each <strong>in</strong>terval.<br />

4) The time tci required <strong>for</strong> the material at a temperature Ti and at a stress 1.5 Sm(Ti,<br />

0) to accumulate a thermal creep stra<strong>in</strong> of 0.05% is obta<strong>in</strong>ed from the curves <strong>in</strong><br />

A.MAT.4.1 of Appendix A (Negligible thermal creep curve).<br />

5) The effect of creep is negligible if the sum of the N ratios of duration ti to the<br />

maximum time tci is less than 1:<br />

N<br />

å 1<br />

i =<br />

( ti/ tc)<br />

£ 1<br />

B 3100 LOW TEMPERATURE RULES<br />

i<br />

The low-temperature rules are appl<strong>ic</strong>able if the negligible thermal creep test B 3050 is<br />

satisfied.<br />

B 3101 Negligible irradiation-<strong>in</strong>duced creep test<br />

At low temperatures where thermal creep is negligible (B 3050), signif<strong>ic</strong>ant time (or neutron<br />

fluence) dependent stra<strong>in</strong> may still occur due to irradiation-<strong>in</strong>duced creep. These stra<strong>in</strong>s may<br />

relax thermal stresses and alter the stress-stra<strong>in</strong> distribution <strong>in</strong> the component. However, if<br />

the fluence at end-of-life is suff<strong>ic</strong>iently low, the effects of irradiation-<strong>in</strong>duced creep can be<br />

ignored <strong>in</strong> the analysis. For a component or a part of a component, irradiation-<strong>in</strong>duced creep<br />

effects can be neglected over the total operat<strong>in</strong>g period if the condition of the test per<strong>for</strong>med<br />

us<strong>in</strong>g the follow<strong>in</strong>g procedure is met.<br />

1) The total operat<strong>in</strong>g period of the component throughout its life and <strong>for</strong> all load<strong>in</strong>gs<br />

<strong>in</strong>clud<strong>in</strong>g temperature and flux <strong>for</strong> wh<strong>ic</strong>h compliance with levels A and C cr<strong>iter</strong>ia<br />

is required, are taken <strong>in</strong>to account.<br />

2) The total operat<strong>in</strong>g period is divided <strong>in</strong>to N time <strong>in</strong>tervals. For each <strong>in</strong>terval i, of a<br />

duration ti, the maximum th<strong>ic</strong>kness-averaged temperature reached Tmi, the mean<br />

neutron flux Fmi, the mean neutron fluence Ftmi and the correspond<strong>in</strong>g value of<br />

1.5Smi(Tmi, Ftmi) are noted.<br />

3) To take full advantage of this rule, the time <strong>in</strong>tervals must be chosen <strong>in</strong> such a<br />

way that the temperature and the neutron flux change as little as possible<br />

throughout the <strong>in</strong>terval.<br />

4) For each <strong>in</strong>terval i, determ<strong>in</strong>e the neutron fluence Ftc2i necessary to accumulate<br />

an effective <strong>in</strong>-reactor creep stra<strong>in</strong> of 0.05% at an effective stress of 1.5Smi(Tmi,<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 37


ITER G 74 MA 8 01-05-28 W0.2<br />

Ftmi), a temperature of Tmi and a neutron flux of Fmi, from A.MAT.4.3 of<br />

appendix A.<br />

5) Compute the follow<strong>in</strong>g sum of the time ratios <strong>for</strong> all N-<strong>in</strong>tervals.<br />

N<br />

tmi<br />

S = å t<br />

f<br />

f 2<br />

i = 1<br />

c i<br />

6) If the above sum, S, is less than 1, then the effects of irradiation-<strong>in</strong>duced creep<br />

can be neglected and time- (or rate-) dependent stress analysis is not needed.<br />

Otherwise, time (or rate) effects should be <strong>in</strong>cluded us<strong>in</strong>g either elast<strong>ic</strong>irradiation-<strong>in</strong>duced-creep<br />

analysis (B 3024.1.1) or, if plast<strong>ic</strong>ity effects are to be<br />

considered at the same time, detailed elasto-visco-plast<strong>ic</strong> analysis (B 3024.4).<br />

Note: The above rule should be used with caution <strong>for</strong> determ<strong>in</strong><strong>in</strong>g whether irradiation<br />

<strong>in</strong>duced creep can relax bolt preloads signif<strong>ic</strong>antly. For example, 0.05% creep stra<strong>in</strong><br />

can cause a bolt stress relaxation of ~ 100 MPa <strong>in</strong> a material with Young's modulus<br />

200 GPa.<br />

B 3200 RULES FOR THE PREVENTION OF M TYPE DAMAGE<br />

B 3211 Immediate plast<strong>ic</strong> collapse and plast<strong>ic</strong> <strong>in</strong>stability<br />

B 3211.1 Elast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> collapse and plast<strong>ic</strong><br />

<strong>in</strong>stability)<br />

Eqs (1), (2), and (3) of IC 3211.1 are <strong>in</strong>tended to guard aga<strong>in</strong>st immediate plast<strong>ic</strong> collapse by<br />

general primary membrane (IC 2522) plus bend<strong>in</strong>g (IC 2523) stresses and by local primary<br />

membrane stress (IC 2524), respectively. Note that the allowable stress Sm (IC 2723) has to<br />

be evaluated at th<strong>ic</strong>kness-averaged temperature and neutron fluence and obta<strong>in</strong>ed from<br />

Appendix A.<br />

B 3211.1.1 Bend<strong>in</strong>g shape factor<br />

The effective bend<strong>in</strong>g shape factor, Keff <strong>in</strong> Eq. 2 of ICÊ3211.1, depends on several variables,<br />

<strong>in</strong>clud<strong>in</strong>g the amount of work-harden<strong>in</strong>g <strong>in</strong> the material, the stra<strong>in</strong> to failure, the def<strong>in</strong>ition of<br />

Sm, and the safety factor that is desired. A detailed discussion of these variables is given <strong>in</strong><br />

Appendix C, section C 3211.1.1. The conclusion of this discussion is that <strong>for</strong> a material with<br />

ample ductility, as is normally the case with unirradiated <strong>structural</strong> materials, then Keff is<br />

given by<br />

Keff = K (2a)<br />

where K is the ratio of the maximum bend<strong>in</strong>g moment achievable, assum<strong>in</strong>g rigid-perfectlyplast<strong>ic</strong><br />

material, to the bend<strong>in</strong>g moment at <strong>in</strong>itial yield, assum<strong>in</strong>g l<strong>in</strong>ear elast<strong>ic</strong> material with<br />

the same yield stress. Values of K <strong>for</strong> simple section geometries are given <strong>in</strong> Figures IC<br />

3211-1, -2, and -3.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 38


ITER G 74 MA 8 01-05-28 W0.2<br />

For a material with reduced ductility, such as irradiated material, then Keff may be<br />

approximated as<br />

( )<br />

( ) -<br />

K = 1 + 2 K -1<br />

K 1<br />

eff eff × rect<br />

where<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 39<br />

(2b)<br />

K is the rigid-perfectly-plast<strong>ic</strong> bend<strong>in</strong>g shape factor <strong>for</strong> the section<br />

under consideration and<br />

KeffÊrect is the effective bend<strong>in</strong>g shape factor <strong>for</strong> a rectangular cross section,<br />

tabulated as a function of neutron fluence and temperature <strong>for</strong> each<br />

material <strong>in</strong> A.MAT.5.2 of Appendix A. The equations used to<br />

derive the tabulated values are given <strong>in</strong> Appendix C, section C<br />

3211.1.1.<br />

For a cross-section consist<strong>in</strong>g of two plates separated by a distance comparable to the plate<br />

th<strong>ic</strong>knesses, as <strong>in</strong> the ITER first wall, the K-value can be obta<strong>in</strong>ed from Figure IC 3211-3.<br />

For a cross section consist<strong>in</strong>g of two th<strong>in</strong> sk<strong>in</strong>s separated by large distance, as <strong>in</strong> the ITER<br />

vacuum <strong>vessel</strong> or cryostat, the theoret<strong>ic</strong>al value of K is 1.0, and Keff = 1.0 regardless of the<br />

material properties. Thus, treat<strong>in</strong>g such a structure as a composite shape <strong>in</strong> bend<strong>in</strong>g is<br />

equivalent to treat<strong>in</strong>g each sk<strong>in</strong> <strong>in</strong>dependently as a membrane. However, if there are local<br />

bend<strong>in</strong>g effects <strong>in</strong> a sk<strong>in</strong>, such as might be caused by pressure load<strong>in</strong>g <strong>in</strong>side the<br />

corrugations, then separate analysis of the <strong>in</strong>dividual sk<strong>in</strong> is required. In this case, each<br />

separate sk<strong>in</strong> then becomes a plate with rectangular cross section and K = 1.5. This<br />

illustrates the po<strong>in</strong>t that the value of K depends on how the structure is idealized. In general,<br />

the shape used <strong>for</strong> K must be consistent with the cross section upon wh<strong>ic</strong>h the stresses are<br />

<strong>in</strong>tegrated to obta<strong>in</strong> the bend<strong>in</strong>g moment (or stress).<br />

B 3211.2 Elasto-plast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> collapse)<br />

If irradiation-<strong>in</strong>duced swell<strong>in</strong>g is not negligible (B 3022), then swell<strong>in</strong>g together with<br />

irradiation-<strong>in</strong>duced creep have to be <strong>in</strong>cluded <strong>in</strong> the stress analysis us<strong>in</strong>g either the simplified<br />

<strong>in</strong>elast<strong>ic</strong> method (B 3024.1.1.1 and B 3024.2) or elasto-visco-plast<strong>ic</strong> analysis (B 3024.4).<br />

However, the rules of IC 3211.2 should be applied us<strong>in</strong>g only the plast<strong>ic</strong> stra<strong>in</strong>s, exclud<strong>in</strong>g<br />

the irradiation-<strong>in</strong>duced creep stra<strong>in</strong> <strong>components</strong>, wh<strong>ic</strong>h are commonly held to be nondamag<strong>in</strong>g.<br />

B 3212 Immediate plast<strong>ic</strong> flow localization<br />

B 3212.1 Elast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> flow localization)<br />

Values of Se as calculated by Eq. (1) are tabulated <strong>in</strong> Table A.MAT.5.3 of appendix A. Eq.<br />

(1) of IC 3212.1 is <strong>in</strong>tended to guard aga<strong>in</strong>st plast<strong>ic</strong> <strong>in</strong>stability (neck<strong>in</strong>g) and plast<strong>ic</strong> flow<br />

localization due to a high primary plus secondary membrane load<strong>in</strong>g. A conservative way of<br />

satisfy<strong>in</strong>g this requirement might be to consider all membrane stresses, whatever the source<br />

(i.e., primary or secondary), as primary. In that case, Eqs.(1), Eq. (2), Eq.(3), or Eq. (4) of IC<br />

3211.1 becomes controll<strong>in</strong>g and Eq. (1) of IC 3212.1 becomes redundant. However, if the<br />

secondary stress is not expected to have a large elast<strong>ic</strong> follow-up, then Eq.(1) may be used


ITER G 74 MA 8 01-05-28 W0.2<br />

with an elast<strong>ic</strong> follow-up factor r as def<strong>in</strong>ed <strong>in</strong> IC 2724. Note that when the uni<strong>for</strong>m<br />

elongation of the material drops below 2%, r1 is set equal to <strong>in</strong>f<strong>in</strong>ity, wh<strong>ic</strong>h effectively<br />

implies that the secondary stresses are considered as primary.<br />

If swell<strong>in</strong>g is not negligible (B 3022), then swell<strong>in</strong>g-<strong>in</strong>duced stresses (B 3024.1.1.1) have to<br />

be <strong>in</strong>cluded with the secondary stresses.<br />

In determ<strong>in</strong><strong>in</strong>g the allowable stress Se, both Su,m<strong>in</strong> and eu have to be evaluated at the same<br />

th<strong>ic</strong>kness-averaged fluence and temperature.<br />

In general, this rule does not become controll<strong>in</strong>g until the material has lost signif<strong>ic</strong>ant<br />

uni<strong>for</strong>m elongation due to irradiation. Up until then, this rule need not be satisfied, as<br />

<strong>in</strong>d<strong>ic</strong>ated by the entry "no limit" <strong>in</strong> Table A.MAT.5.3.<br />

Note: The value of r1 <strong>for</strong> Eq. (1) has been chosen conservatively to be equal to 4 (IC 2724).<br />

In some cases r1 may be shown to be much less than 4. The <strong>design</strong>er may choose to reduce<br />

the conservativeness by us<strong>in</strong>g a smaller value of r1 if it can be justified us<strong>in</strong>g the procedure<br />

given <strong>in</strong> B 3024.2.<br />

B 3212.2 Elasto-plast<strong>ic</strong> analysis (Immediate plast<strong>ic</strong> flow localization)<br />

If irradiation-<strong>in</strong>duced swell<strong>in</strong>g is not negligible (B 3022), then swell<strong>in</strong>g together with<br />

irradiation-<strong>in</strong>duced creep have to be <strong>in</strong>cluded <strong>in</strong> the stress analysis us<strong>in</strong>g either the simplified<br />

<strong>in</strong>elast<strong>ic</strong> method (B 3024.1.1.1 and B 3024.2) or elasto-visco-plast<strong>ic</strong> analysis (B 3024.4).<br />

However, the stra<strong>in</strong> limits of IC 3212.2 should be applied to the plast<strong>ic</strong> stra<strong>in</strong>, exclud<strong>in</strong>g the<br />

irradiation-<strong>in</strong>duced creep stra<strong>in</strong> <strong>components</strong>.<br />

B 3213 Immediate local fracture due to exhaustion of ductility<br />

B 3213.1 Elast<strong>ic</strong> analysis (Immediate local fracture due to exhaustion<br />

of ductility)<br />

Equation (1) of IC 3213.1 has been provided to guard aga<strong>in</strong>st crack<strong>in</strong>g <strong>in</strong> regions of stress<br />

concentration <strong>in</strong>clud<strong>in</strong>g peak stress effects while Eq. (2) is <strong>for</strong> guard<strong>in</strong>g aga<strong>in</strong>st crack<strong>in</strong>g<br />

exclud<strong>in</strong>g peak stress effects, e.g., crack<strong>in</strong>g of extreme fibers <strong>in</strong> bend<strong>in</strong>g due to exhaustion of<br />

ductility by irradiation. The allowable stresses Sd (T, ft, r2) and Sd (T, ft, r3) <strong>for</strong> the two<br />

cases are given <strong>in</strong> A.MAT.5.4 of appendix A. The value of the elast<strong>ic</strong> follow-up factor r2 has<br />

been set on the basis of analysis of notched tensile tests (Appendix C 3024.1.4) and on the<br />

assumption that KT = 4. The value of r3 has been chosen conservatively on the basis of<br />

analysis of three-po<strong>in</strong>t bend tests (Appendix C 3024.1.4). As be<strong>for</strong>e, the <strong>design</strong>er may<br />

choose to reduce the conservativeness by us<strong>in</strong>g a smaller value of r2 or r3 if it can be justified<br />

us<strong>in</strong>g the procedure given <strong>in</strong> B 3024.2.<br />

If swell<strong>in</strong>g is not negligible (B 3022), then swell<strong>in</strong>g-<strong>in</strong>duced stresses (B 3024.1.1.1) have to<br />

be <strong>in</strong>cluded with the secondary stresses.<br />

Note 1: For materials like 316L(N) sta<strong>in</strong>less steel, the true stra<strong>in</strong> at rupture rema<strong>in</strong>s high at<br />

temperatures £ 400¡C and does not reduce signif<strong>ic</strong>antly up to a fluence of 10 dpa. For these<br />

materials, the PL+Pb+Q+F limit need not be checked because, crack<strong>in</strong>g at notch roots due to<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 40


ITER G 74 MA 8 01-05-28 W0.2<br />

exhaustion of ductility is not a problem (the elast<strong>ic</strong> follow up factor r2 rema<strong>in</strong>s bounded by<br />

the stress concentration factor, irrespective of stra<strong>in</strong> harden<strong>in</strong>g capability).<br />

Note 2: For materials that suffer severe loss of stra<strong>in</strong> harden<strong>in</strong>g capability (i.e., uni<strong>for</strong>m<br />

elongation) due to irradiation, the elast<strong>ic</strong> follow up factor r3 can become very large <strong>in</strong> areas of<br />

high elast<strong>ic</strong> follow up and the Sd limit <strong>for</strong> PL+Pb+Q has to be checked. However, this rule<br />

does not become controll<strong>in</strong>g until the material has lost signif<strong>ic</strong>ant stra<strong>in</strong> harden<strong>in</strong>g capability<br />

due to irradiation. Until then, this rule need not be checked, as <strong>in</strong>d<strong>ic</strong>ated by the entry "no<br />

limit" <strong>in</strong> Table A.MAT.5.4.<br />

B 3213.2 Elasto-plast<strong>ic</strong> analysis (Immediate local fracture due to<br />

exhaustion of ductility)<br />

If irradiation-<strong>in</strong>duced swell<strong>in</strong>g is not negligible (B 3022), then swell<strong>in</strong>g together with<br />

irradiation-<strong>in</strong>duced creep have to be <strong>in</strong>cluded <strong>in</strong> the stress analysis us<strong>in</strong>g either the simplified<br />

<strong>in</strong>elast<strong>ic</strong> method (B 3024.1.1.1 and B 3024.2) or elasto-visco-plast<strong>ic</strong> analysis (B 3024.4).<br />

However, the stra<strong>in</strong> limit of IC 3213.2 should be applied only to the plast<strong>ic</strong> stra<strong>in</strong>, exclud<strong>in</strong>g<br />

the irradiation-<strong>in</strong>duced creep stra<strong>in</strong>.<br />

BÊ3214 Fast fracture<br />

B 3214.1 Elast<strong>ic</strong> analysis (Fast fracture)<br />

For the elast<strong>ic</strong> analysis rules to be str<strong>ic</strong>tly appl<strong>ic</strong>able, stresses everywhere have to rema<strong>in</strong><br />

below yield. However, some local yield<strong>in</strong>g may be permitted at the crack tip provided the<br />

yielded zone is small compared to the crack length and the rema<strong>in</strong><strong>in</strong>g ligament th<strong>ic</strong>kness.<br />

Yield<strong>in</strong>g is also permitted if the evaluation is carried out <strong>for</strong> a section suff<strong>ic</strong>iently away from<br />

the yielded region. In all cases, a mode I stress <strong>in</strong>tensity factor KI has to be calculated <strong>for</strong> a<br />

postulated surface crack of depth ao and m<strong>in</strong>imum length 10ao subjected to the <strong>in</strong>d<strong>ic</strong>ated<br />

load<strong>in</strong>g. The value of ao is given by<br />

ao = Max[4au , h/4]<br />

where au = largest undetectable crack by the NDE technique to be used<br />

and h = th<strong>ic</strong>kness of section<br />

For the global fast fracture case, the postulated crack should be oriented normal to the<br />

maximum tensile component of the membrane stress due to all primary (PL) and primary plus<br />

secondary load<strong>in</strong>gs (PL+QL). The stress <strong>in</strong>tensity factor may be obta<strong>in</strong>ed either from a<br />

recognized handbook of stress <strong>in</strong>tensity factors, or elast<strong>ic</strong> f<strong>in</strong>ite element analysis.<br />

For the local fast fracture case, the crack should be oriented normal to the maximum tensile<br />

component of the local stress due to all primary and secondary load<strong>in</strong>gs, <strong>in</strong>clud<strong>in</strong>g peak,<br />

(PL+Pb+Q+F). The stress <strong>in</strong>tensity factor should be calculated either by elast<strong>ic</strong> f<strong>in</strong>ite element<br />

analysis us<strong>in</strong>g crack tip s<strong>in</strong>gular elements or by any other method that can be justified to give<br />

a conservative result.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 41


ITER G 74 MA 8 01-05-28 W0.2<br />

Note: A conservative estimate <strong>for</strong> KI can be obta<strong>in</strong>ed by us<strong>in</strong>g the follow<strong>in</strong>g expression <strong>for</strong> a<br />

s<strong>in</strong>gle-edge notched plate subjected to a uni<strong>for</strong>m tensile stress s at <strong>in</strong>f<strong>in</strong>ity<br />

K<br />

I<br />

1. 122<br />

=<br />

æ ao<br />

1 -<br />

ö<br />

è h ø<br />

3/ 2 s p<br />

a<br />

B 3214.2 Elasto-plast<strong>ic</strong> analysis (Fast fracture)<br />

o<br />

If the postulated crack <strong>for</strong> fast fracture analysis is embedded <strong>in</strong> a yielded region, l<strong>in</strong>ear elast<strong>ic</strong><br />

fracture mechan<strong>ic</strong>s methodology may give highly unconservative results. In such cases,<br />

elasto-plast<strong>ic</strong> fracture mechan<strong>ic</strong>s methodology should be used. Any procedure based on<br />

elasto-plast<strong>ic</strong> fracture mechan<strong>ic</strong>s (such as section XI of ASME, Appendix A16 of RCC-MR,<br />

R6 rule) may be considered but its appl<strong>ic</strong>ation shall be justified. Special attention shall be<br />

given to the effect of thermal loads, complex load<strong>in</strong>g modes, and spread<strong>in</strong>g of plast<strong>ic</strong>ity. The<br />

J-<strong>in</strong>tegral is an acceptable cr<strong>iter</strong>ion <strong>for</strong> such cases. Most commercially available f<strong>in</strong>ite<br />

element programs (e.g., ANSYS, ABAQUS, etc.) have the capability to compute the J<strong>in</strong>tegral<br />

either by a contour <strong>in</strong>tegral method or by an energy method.<br />

B 3300 RULES FOR THE PREVENTION OF C TYPE DAMAGE (LEVELS A<br />

AND C)<br />

B 3310 Progressive de<strong>for</strong>mation or ratchet<strong>in</strong>g<br />

B 3311 Elast<strong>ic</strong> analysis (Progressive de<strong>for</strong>mation or ratchet<strong>in</strong>g)<br />

B 3311.1 3Sm rule<br />

The 3Sm limit <strong>for</strong> the PL+Pb+D(P+Q) stresses ensures that the membrane and bend<strong>in</strong>g<br />

stresses will shakedown after the first cycles of load<strong>in</strong>g and unload<strong>in</strong>g, thus ensur<strong>in</strong>g that<br />

there will be no overall (large scale) progressive de<strong>for</strong>mation. Only peak stresses (F), wh<strong>ic</strong>h<br />

are local <strong>in</strong> nature, may cont<strong>in</strong>ue to cycle plast<strong>ic</strong>ally. Peak stresses are considered <strong>in</strong> the<br />

per<strong>for</strong>mance of fatigue evaluation and also <strong>for</strong> fracture <strong>in</strong> embrittled materials. Indeed, the<br />

requirement that there be no overall progressive de<strong>for</strong>mation is necessary to ensure that the<br />

fatigue tests, wh<strong>ic</strong>h are conducted without progressive de<strong>for</strong>mation, represent the behaviour<br />

of the structure.<br />

All cycl<strong>ic</strong> stress ranges <strong>in</strong>clud<strong>in</strong>g the stress range due to disruption-<strong>in</strong>duced electromagnet<strong>ic</strong><br />

load<strong>in</strong>g have to be added to the secondary stress range (see appendix C 3311.3). For annealed<br />

austenit<strong>ic</strong> sta<strong>in</strong>less steels, if 0.9Sy controls the Sm value, some ratchet<strong>in</strong>g stra<strong>in</strong> may occur<br />

until stra<strong>in</strong> harden<strong>in</strong>g <strong>in</strong>creases the cycl<strong>ic</strong> yield stress suff<strong>ic</strong>iently to achieve shakedown.<br />

B 3311.2 Bree-diagram rule<br />

S<strong>in</strong>ce the Bree diagram approach is mathemat<strong>ic</strong>ally equivalent to the Eff<strong>ic</strong>iency diagram<br />

approach of RCC-MR (see appendix C 3311), the Bree diagram rule was adopted <strong>for</strong> the<br />

SDC-IC. However, the expression <strong>for</strong> the Bree diagram used <strong>in</strong> the SDC-IC is slightly<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 42


ITER G 74 MA 8 01-05-28 W0.2<br />

different from that <strong>in</strong> the ASME Code, Section III. In the SDC-IC, the normaliz<strong>in</strong>g factor <strong>for</strong><br />

the stresses is the average of the Sy values at the maximum and the m<strong>in</strong>imum wall-averaged<br />

temperature and neutron fluence dur<strong>in</strong>g the cycle be<strong>in</strong>g evaluated <strong>for</strong> all values of Sm, In the<br />

ASME Code, it is permissible to use 1.5ÊSm , <strong>in</strong>stead of Sy , whenever it is greater than Sy.<br />

The reason <strong>for</strong> choos<strong>in</strong>g the Sy value <strong>in</strong> this way is to m<strong>in</strong>imize the ratchet<strong>in</strong>g stra<strong>in</strong><br />

accumulated prior to shakedown and to assure that the allowable stresses determ<strong>in</strong>ed by the<br />

Bree diagram rule fall with<strong>in</strong> (i.e., are conservative with respect to) those determ<strong>in</strong>ed by the<br />

Eff<strong>ic</strong>iency diagram rule of RCC-MR (see Appendix C 3311).<br />

The Bree diagram approach is less conservative than the 3Sm rule. However, s<strong>in</strong>ce the yield<br />

stress rather than the proportional limit is used <strong>in</strong> the analysis, a small amount of ratchet<strong>in</strong>g<br />

stra<strong>in</strong> (~ 0.2%) may occur until stra<strong>in</strong> harden<strong>in</strong>g raises the proportional limit to the yield<br />

stress. If this is unacceptable, the 3Sm rule should be used <strong>in</strong>stead.<br />

B 3312 Elasto-plast<strong>ic</strong> analysis (Progressive de<strong>for</strong>mation or<br />

ratchet<strong>in</strong>g)<br />

Rules <strong>for</strong> prevent<strong>in</strong>g progressive de<strong>for</strong>mation based on elast<strong>ic</strong> analysis are conta<strong>in</strong>ed <strong>in</strong><br />

ICÊ3311. They require the partition of stresses <strong>in</strong>to primary and secondary <strong>components</strong>. In<br />

certa<strong>in</strong> cases, this partition can lead to overly conservative results with regard to the<br />

ratchet<strong>in</strong>g resistance of the component. Elasto-plast<strong>ic</strong> analysis could provide a less<br />

conservative evaluation.<br />

In order to check the limits on these stra<strong>in</strong>s, an elasto-plast<strong>ic</strong> analysis, giv<strong>in</strong>g either the exact<br />

value or an upper bound to the stra<strong>in</strong>s result<strong>in</strong>g from all the cycles envisaged, should be<br />

carried out. This can be obta<strong>in</strong>ed:<br />

- either by a cycl<strong>ic</strong> elasto-plast<strong>ic</strong> calculation until stabilization of stress-stra<strong>in</strong><br />

response, or<br />

- by extrapolated cycl<strong>ic</strong> elasto-plast<strong>ic</strong> calculation, us<strong>in</strong>g a validated method.<br />

If a certa<strong>in</strong> block of cycles (conta<strong>in</strong><strong>in</strong>g one or more types of cycles) is repeated period<strong>ic</strong>ally,<br />

the block can be represented by one type of cycle (called the envelope cycle) wh<strong>ic</strong>h<br />

sequentially adds all cycles <strong>in</strong> the block. (The <strong>in</strong>dividual cycles <strong>for</strong> wh<strong>ic</strong>h the l<strong>in</strong>ear stress<br />

does not <strong>in</strong>duce plast<strong>ic</strong> stra<strong>in</strong> may be excluded). The analysis can then be simplified as<br />

follows:<br />

- analyze <strong>in</strong> detail all the cycles <strong>in</strong>side one envelope cycle,<br />

- compute a ratchet<strong>in</strong>g stra<strong>in</strong> <strong>in</strong>crement <strong>in</strong> a s<strong>in</strong>gle appl<strong>ic</strong>ation of the envelope<br />

cycle.<br />

- the usage fraction <strong>for</strong> each type of envelope cycle is this ratchet<strong>in</strong>g stra<strong>in</strong><br />

<strong>in</strong>crement multiplied by the number of times the envelope cycle is repeated,<br />

divided by the allowable ductility term.<br />

For the envelope cycle elasto-plast<strong>ic</strong> analysis, and <strong>in</strong> the absence of expl<strong>ic</strong>it <strong>in</strong>structions, an<br />

<strong>in</strong>itial zero stress and stra<strong>in</strong> condition may be assumed. However, the stress and stra<strong>in</strong><br />

conditions at the beg<strong>in</strong>n<strong>in</strong>g of each cycle conta<strong>in</strong>ed with<strong>in</strong> the envelope cycle are those<br />

rema<strong>in</strong><strong>in</strong>g at the end of the calculation of the previous cycle.<br />

The permissible limits <strong>for</strong> the accumulated membrane and peak stra<strong>in</strong>s are given <strong>in</strong> IC 3312.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 43


ITER G 74 MA 8 01-05-28 W0.2<br />

B 3320 Time-<strong>in</strong>dependent fatigue<br />

B 3322 Limits on fatigue damage<br />

B 3322.1 Calculation of the fatigue usage fraction: V( De)<br />

Follow<strong>in</strong>g the estimation of the "real" stra<strong>in</strong> ranges De (B 3323) of each cycle, the fatigue<br />

usage fraction V ( De)<br />

at the po<strong>in</strong>t under consideration is determ<strong>in</strong>ed <strong>in</strong> accordance with B<br />

2752, us<strong>in</strong>g the fatigue curves <strong>in</strong> A---- <strong>for</strong> all load<strong>in</strong>g cycles requir<strong>in</strong>g compliance with levels<br />

A and C cr<strong>iter</strong>ia.<br />

B 3322.2 Estimation of irradiation effects on fatigue usage fraction<br />

If fatigue curves <strong>for</strong> irradiated material are not available, the fatigue usage fraction may be<br />

estimated by us<strong>in</strong>g an empir<strong>ic</strong>al <strong>for</strong>mula <strong>for</strong> estimat<strong>in</strong>g the fatigue curve (described below).<br />

Follow<strong>in</strong>g is a description of the general procedure <strong>for</strong> calculat<strong>in</strong>g the fatigue usage fraction,<br />

followed by the detail of how the fatigue curve can be estimated.<br />

BÊ3322.2.1 Calculation of fatigue usage fraction - general<br />

- The complete load<strong>in</strong>g history is considered to have N types of cycles.<br />

- For each type of cycle j (j=1 to N), the total operat<strong>in</strong>g period is divided <strong>in</strong>to Mj<br />

time <strong>in</strong>tervals, dur<strong>in</strong>g each of wh<strong>ic</strong>h the temperature and fluence are assumed to<br />

be constant.<br />

- For each time <strong>in</strong>terval k (k= 1 to Mj ), of duration tk, there are nk cycles of type j.<br />

- The total number of cycles of type j dur<strong>in</strong>g the complete load<strong>in</strong>g history is then<br />

M j<br />

å 1<br />

n = n<br />

j k<br />

k =<br />

- The maximum temperature reached <strong>in</strong> time <strong>in</strong>terval k (k= 1 to Mj ) is denoted by<br />

Tk and the mean neutron flux is denoted by Fk.<br />

- The maximum values of the temperature and neutron fluence reached dur<strong>in</strong>g the<br />

period under consideration should generally be used <strong>in</strong> the follow<strong>in</strong>g estimation<br />

procedures. However, s<strong>in</strong>ce the irradiated material may display a ductilitym<strong>in</strong>imum<br />

as a function of temperature, care should be taken to ensure that the<br />

fatigue damage is estimated conservatively.<br />

- To take full advantage of this rule, the <strong>in</strong>tervals of time must be chosen <strong>in</strong> such a<br />

way that the neutron flux changes as little as possible throughout each <strong>in</strong>terval.<br />

- The cumulative fatigue usage fraction is the sum of the fatigue usage fractions<br />

calculated <strong>for</strong> all types of stra<strong>in</strong> cycles:<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 44


ITER G 74 MA 8 01-05-28 W0.2<br />

V<br />

=<br />

j=<br />

1<br />

é<br />

ê<br />

ê<br />

ë<br />

M<br />

N j<br />

å å<br />

1<br />

k =<br />

n<br />

N<br />

k<br />

( )<br />

j k<br />

ù<br />

ú<br />

ú<br />

û<br />

BÊ3322.2.2 Procedure <strong>for</strong> estimat<strong>in</strong>g the fatigue curve<br />

- For each type of stra<strong>in</strong> cycle j, the <strong>design</strong> allowable number of cycles ( N j ) k<br />

correspond<strong>in</strong>g to the temperature Tk and neutron fluence conditions of the <strong>in</strong>terval<br />

of time k, and correspond<strong>in</strong>g to a stra<strong>in</strong> range De, can be estimated by the<br />

follow<strong>in</strong>g general equation:<br />

D F é<br />

ù F é<br />

e BNj AN<br />

ëê k<br />

ëê<br />

e = ( )<br />

b<br />

c<br />

p j<br />

k<br />

ûú + ( )<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 45<br />

ù<br />

ûú<br />

where, denot<strong>in</strong>g the unirradiated values by the subscripts u and irradiated<br />

values correspond<strong>in</strong>g to the neutron fluence Ftk , by the subscripts i:<br />

F e = elast<strong>ic</strong> stra<strong>in</strong>-life modif<strong>ic</strong>ation factor (£ 1)<br />

= the ratio between the m<strong>in</strong>imum irradiated ultimate tensile<br />

strength, wh<strong>ic</strong>h is a function of temperature and neutron<br />

fluence and the m<strong>in</strong>imum ultimate tensile strength, wh<strong>ic</strong>h is a<br />

function of temperature and <strong>in</strong>dependent of neutron fluence:<br />

=<br />

( S )<br />

( S )<br />

u i<br />

u u<br />

, <strong>for</strong> an irradiation harden<strong>in</strong>g material,<br />

F p = plast<strong>ic</strong> stra<strong>in</strong>-life modif<strong>ic</strong>ation factor (£ 1)<br />

where<br />

fk<br />

( )<br />

= f k<br />

=<br />

e<br />

e<br />

tru<br />

tri<br />

c<br />

= the ratio between the m<strong>in</strong>imum unirradiated ductility<br />

(etruÊ=Ê etr ( T,0 k )), wh<strong>ic</strong>h is a function of temperature and<br />

<strong>in</strong>dependent of fluence and the m<strong>in</strong>imum irradiated ductility<br />

(etriÊ=Ê etr ( Tk, Ftk<br />

)), wh<strong>ic</strong>h is a function of temperature and<br />

fluence,<br />

A and B are the coeff<strong>ic</strong>ients and b and c are the exponents that fit the <strong>design</strong><br />

fatigue curve (A1.5.4) <strong>for</strong> the unirradiated material, <strong>in</strong> the <strong>for</strong>m of the<br />

follow<strong>in</strong>g equation:<br />

b c<br />

De = BN + AN


ITER G 74 MA 8 01-05-28 W0.2<br />

A conservative approximation of the fatigue usage fraction can be obta<strong>in</strong>ed by<br />

consider<strong>in</strong>g Fe = 0 . In that case, the fatigue usage fraction <strong>for</strong> the type of stra<strong>in</strong><br />

cycle j is equal to the sum of the fatigue usage fraction <strong>for</strong> each <strong>in</strong>terval k, wh<strong>ic</strong>h<br />

is the ratio of the number of stra<strong>in</strong> cycles nk to the maximum allowable number<br />

( N j ) <strong>for</strong> this type of cycle, determ<strong>in</strong>ed from the unirradiated fatigue curves,<br />

k<br />

multiplied by the correction factor f k:<br />

N é M j<br />

V = å êå<br />

f<br />

j=<br />

1 ê<br />

ëk<br />

= 1<br />

k<br />

n<br />

N<br />

k<br />

( )<br />

j k<br />

B 3323. Calculation of equivalent stra<strong>in</strong> range De<br />

ù<br />

ú<br />

ú<br />

û<br />

B 3323.1 Elast<strong>ic</strong> analysis (Time-<strong>in</strong>dependent fatigue)<br />

When elast<strong>ic</strong> analysis is used to calculate the response of a structure, the range of stra<strong>in</strong>s<br />

obta<strong>in</strong>ed does not account <strong>for</strong> plast<strong>ic</strong> stra<strong>in</strong>s wh<strong>ic</strong>h would occur if the real behavior of the<br />

material were taken <strong>in</strong>to consideration . The method outl<strong>in</strong>ed below is aimed at provid<strong>in</strong>g an<br />

estimate of the "real" stra<strong>in</strong> range De on the basis of the results of the elast<strong>ic</strong> analysis. This is<br />

achieved by evaluat<strong>in</strong>g the amplif<strong>ic</strong>ation of the stra<strong>in</strong>, and the result<strong>in</strong>g stra<strong>in</strong> range, due to<br />

plast<strong>ic</strong>ity as well as cycl<strong>ic</strong> harden<strong>in</strong>g or soften<strong>in</strong>g of the material as represented by the cycl<strong>ic</strong><br />

stress-stra<strong>in</strong> curves given <strong>in</strong> A.5.9.<br />

( ( ) ) at<br />

To apply the rules of this section, the total stress <strong>in</strong>tensity range Dstot = D P + Q + F<br />

the po<strong>in</strong>t under consideration and <strong>for</strong> each of the cycles must be calculated elast<strong>ic</strong>ally, us<strong>in</strong>g<br />

the procedure given <strong>in</strong> B 2550. This range can be obta<strong>in</strong>ed either by a suff<strong>ic</strong>iently detailed<br />

calculation of the region concerned or by us<strong>in</strong>g an appropriate stress concentration factor.<br />

The value of the total effective stra<strong>in</strong> range De is the sum of four scalar stra<strong>in</strong>s<br />

De , De , De , De<br />

:<br />

1 2 3 4<br />

De = De + De + De + De<br />

1 2 3 4<br />

These four terms are determ<strong>in</strong>ed us<strong>in</strong>g a uniaxial cycl<strong>ic</strong> stress-stra<strong>in</strong> curve (see A.MAT.5.7 <strong>in</strong><br />

Appendix A) correspond<strong>in</strong>g to the highest temperature (Tmax) and the lowest neutron fluence<br />

(Ft)m<strong>in</strong> at the po<strong>in</strong>t exam<strong>in</strong>ed dur<strong>in</strong>g the cycle concerned.<br />

For cycl<strong>ic</strong>ally harden<strong>in</strong>g materials, the monoton<strong>ic</strong> stress-stra<strong>in</strong> curve is taken as a lower<br />

bound of the cycl<strong>ic</strong> curve.<br />

For irradiation harden<strong>in</strong>g materials, the virg<strong>in</strong> material cycl<strong>ic</strong> curve is taken as a lower bound<br />

of the irradiated material cycl<strong>ic</strong> curve.<br />

Calculation of De 1:<br />

B 3323.1.1 Elast<strong>ic</strong> stra<strong>in</strong> range<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 46


ITER G 74 MA 8 01-05-28 W0.2<br />

D e 1 represents the stra<strong>in</strong> range given by elast<strong>ic</strong> analysis (path a-b <strong>in</strong> Diagram 1,<br />

FigureÊBÊ3323-1).<br />

Ds tot<br />

Diagram 1<br />

Ds<br />

De 1<br />

a b<br />

Cycl<strong>ic</strong> stress-stra<strong>in</strong><br />

Curve<br />

Figure B 3323-1: Determ<strong>in</strong>ation of total stra<strong>in</strong> range <strong>for</strong> fatigue<br />

us<strong>in</strong>g elast<strong>ic</strong> analysis. Step 1 - De1<br />

D e 1 may be calculated on the basis of elast<strong>ic</strong> analysis <strong>in</strong> accordance with the rules def<strong>in</strong>ed <strong>in</strong><br />

B 2630, or alternatively, the follow<strong>in</strong>g <strong>for</strong>mula may be used:<br />

2<br />

De1n Ds<br />

3 1<br />

= +<br />

where<br />

( ) ( tot E)<br />

E = Young's modulus (A.MAT.2.2) at maximum temperature<br />

Tmax and m<strong>in</strong>imum fluence Ê(Ft)m<strong>in</strong> dur<strong>in</strong>g the cycle,<br />

n = Poisson's ratio (A.MAT.2.3).<br />

B 3323.1.2 Corrections <strong>for</strong> effects of plast<strong>ic</strong>ity<br />

Calculation of De 2 (Cycl<strong>ic</strong> Primary Stress):<br />

De2 represents the plast<strong>ic</strong> stra<strong>in</strong> range due to cycl<strong>ic</strong> primary stress (path b-c <strong>in</strong> Diagram 2,<br />

FigureÊBÊ3323-2).<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 47<br />

De


ITER G 74 MA 8 01-05-28 W0.2<br />

Diagram 2<br />

Ds tot<br />

D[P m +0.67(P b +P L -P m )]<br />

Ds<br />

a b<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 48<br />

De 2<br />

Figure B 3323-2: Determ<strong>in</strong>ation of total stra<strong>in</strong> range <strong>for</strong> fatigue<br />

us<strong>in</strong>g elast<strong>ic</strong> analysis. Step 2 - De2<br />

It can be determ<strong>in</strong>ed us<strong>in</strong>g the follow<strong>in</strong>g procedure:<br />

- us<strong>in</strong>g the procedure outl<strong>in</strong>ed <strong>in</strong> BÊ2550, calculate the effective range of primary<br />

stress,<br />

[ ( ) ]<br />

DP = D P + 067 . P + P - P<br />

eff m b L m<br />

- calculate a plast<strong>ic</strong> stra<strong>in</strong> range De2 correspond<strong>in</strong>g to the effective range of<br />

primary stress as follows:<br />

· us<strong>in</strong>g the cycl<strong>ic</strong> stress stra<strong>in</strong> curves given <strong>in</strong> A.MAT.5.7, calculate the<br />

total cycl<strong>ic</strong> stra<strong>in</strong> range Decycl<strong>ic</strong> correspond<strong>in</strong>g to the effective range<br />

of primary stress D Peff .,<br />

· calculate the plast<strong>ic</strong> part of the total cycl<strong>ic</strong> stra<strong>in</strong> range as<br />

De2 = D e<br />

- DP<br />

E<br />

cycl<strong>ic</strong> eff<br />

The value of De2 is generally very low, but can nonetheless be important when an<br />

appreciable elast<strong>ic</strong> follow-up exists.<br />

Calculation of De3:<br />

D e 3 represents the "plast<strong>ic</strong>" <strong>in</strong>crease <strong>in</strong> stra<strong>in</strong>s along path c-d <strong>in</strong> Diagram 3, Figure BÊ3223-3.<br />

c<br />

De


ITER G 74 MA 8 01-05-28 W0.2<br />

Diagram 3<br />

Ds tot<br />

Ds<br />

a b<br />

c<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 49<br />

d<br />

De 3<br />

Ds De = const<br />

Figure B 3323-3: Determ<strong>in</strong>ation of total stra<strong>in</strong> range <strong>for</strong> fatigue<br />

us<strong>in</strong>g elast<strong>ic</strong> analysis. Step 3 - De3<br />

Po<strong>in</strong>t (d) is the <strong>in</strong>tersection po<strong>in</strong>t of the cycl<strong>ic</strong> curve and the hyperbola D s . D e = constant<br />

pass<strong>in</strong>g through po<strong>in</strong>t (c) with coord<strong>in</strong>ates:<br />

{ De1 + De2 ; Dstot<br />

}<br />

D e 3 has been derived from the above hyperbola such that<br />

( )<br />

( ) +<br />

De = K -1De<br />

De<br />

3 e 1 2<br />

The value of K e has been tabulated <strong>in</strong> appendix A as a function of Ds tot<br />

Calculation of De4:<br />

De4 represents the "plast<strong>ic</strong>" <strong>in</strong>crease <strong>in</strong> stra<strong>in</strong> due to triaxiality. De4 is def<strong>in</strong>ed by the<br />

follow<strong>in</strong>g equation:<br />

( )<br />

De = K -1De<br />

4 n 1<br />

The value of the amplif<strong>ic</strong>ation coeff<strong>ic</strong>ient Kn is evaluated, as shown <strong>in</strong> Diagram 4,<br />

FigureÊBÊ3323-4, us<strong>in</strong>g the curves and tables <strong>in</strong> A.MAT.5.7 <strong>for</strong> temperature Tmax.<br />

De


ITER G 74 MA 8 01-05-28 W0.2<br />

Note 1:<br />

Kn<br />

Kn<br />

1<br />

Diagram 4<br />

Dstot<br />

Max (T, ft)<br />

Figure B 3323-4: Determ<strong>in</strong>ation of the Poisson's ratio factor Kn<br />

<strong>for</strong> fatigue us<strong>in</strong>g elast<strong>ic</strong> analysis<br />

This method is str<strong>ic</strong>tly appl<strong>ic</strong>able <strong>in</strong> the case of an equi-biaxial stress field, such as the<br />

sk<strong>in</strong> effect dur<strong>in</strong>g thermal shock. In other cases, the evaluation of De4 given above is<br />

overly conservative . It may there<strong>for</strong>e be reduced by multiply<strong>in</strong>g the value of Kn given <strong>in</strong><br />

Appendix A by the follow<strong>in</strong>g coeff<strong>ic</strong>ient:<br />

Note 2 :<br />

1 + 3 d 2 m 2<br />

( ) 1 + 3 d 2 m 2<br />

( )<br />

where:<br />

d<br />

m<br />

=<br />

=<br />

1 +<br />

1 -<br />

1 -<br />

1 +<br />

n<br />

n<br />

n<br />

n<br />

m = m K n<br />

s - s<br />

.<br />

s + s<br />

D D<br />

D D<br />

1 2<br />

1 2<br />

Ds1 = stress range calculated elast<strong>ic</strong>ally <strong>in</strong> a pr<strong>in</strong>cipal direction.<br />

Ds2 = stress range calculated elast<strong>ic</strong>ally <strong>in</strong> the other pr<strong>in</strong>cipal<br />

direction.<br />

n = Poisson's ratio.<br />

The <strong>in</strong>itial slope of Diagrams 1 to 3 above must <strong>in</strong> reality be equal to 3 E/[2 (1 + n)]. In<br />

fact, this is not the case with the curves given <strong>in</strong> Appendix A1, wh<strong>ic</strong>h are obta<strong>in</strong>ed from<br />

uniaxial tests. In pract<strong>ic</strong>e, this difference is negligible <strong>for</strong> the constructions shown <strong>in</strong><br />

diagrams 2 and 3.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 50<br />

Ds


ITER G 74 MA 8 01-05-28 W0.2<br />

B 3323.1.3 Comb<strong>in</strong><strong>in</strong>g Components<br />

Add<strong>in</strong>g the four stra<strong>in</strong> <strong>components</strong>,<br />

De = De + De + De + De<br />

1 2 3 4<br />

( ) +<br />

( ) + -<br />

( )<br />

De = De + De + K -1<br />

De De K 1 De<br />

1 2 e 1 2 n 1<br />

( ) +<br />

De £ 1 + K - 1 + K -1<br />

De De<br />

e n<br />

( 1 2)<br />

= ( Ke + Kn<br />

-1)<br />

( De1 + De2)<br />

The amplif<strong>ic</strong>ation coeff<strong>ic</strong>ients Ke( Dstot, T, Ft)<br />

and KnDstot, T, Ft<br />

Appendix A.<br />

( ) are tabulated <strong>in</strong><br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 51


ITER G 74 MA 8 01-05-28 W0.2<br />

B 3400 RULES FOR THE PREVENTION OF BUCKLING<br />

The structures considered <strong>in</strong> this Section are th<strong>in</strong> shells, i.e., structures wh<strong>ic</strong>h can be<br />

represented by a mean surface and a th<strong>ic</strong>kness.<br />

The procedure given here <strong>for</strong> <strong>in</strong>stability stress analysis <strong>for</strong> load-controlled buckl<strong>in</strong>g is based<br />

on that <strong>in</strong> RCC-MR. Other methods may be used if they can be justified. RCC-MR does not<br />

require <strong>in</strong>stability analysis <strong>for</strong> stra<strong>in</strong>-controlled buckl<strong>in</strong>g. However, as <strong>in</strong> ASME Code Case<br />

N47, an <strong>in</strong>stability analysis <strong>for</strong> stra<strong>in</strong>-controlled buckl<strong>in</strong>g is required <strong>in</strong> the SDC-IC but with<br />

smaller load factors than those <strong>for</strong> load-controlled buckl<strong>in</strong>g.<br />

Note: The guidel<strong>in</strong>es provided <strong>in</strong> sections B 3410 and B 3420 are appl<strong>ic</strong>able to stat<strong>ic</strong> or<br />

quasi-stat<strong>ic</strong> buckl<strong>in</strong>g without any time or rate-dependent effects. If the load<strong>in</strong>g under<br />

consideration is due to a fast transient, e.g., electromagnet<strong>ic</strong> load<strong>in</strong>g dur<strong>in</strong>g plasma<br />

disruptions, dynam<strong>ic</strong> effects may have a signif<strong>ic</strong>ant <strong>in</strong>fluence on buckl<strong>in</strong>g. Guidel<strong>in</strong>es <strong>for</strong><br />

treat<strong>in</strong>g dynam<strong>ic</strong> buckl<strong>in</strong>g and irradiation-<strong>in</strong>duced creep buckl<strong>in</strong>g will be provided <strong>in</strong> the<br />

future.<br />

B 3420 Buckl<strong>in</strong>g limits<br />

B 3421 Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability under monoton<strong>ic</strong><br />

load<strong>in</strong>g<br />

B 3421.1 Elast<strong>ic</strong> analysis (Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability )<br />

B 3421.1.1 Load-controlled buckl<strong>in</strong>g limits<br />

The method described <strong>in</strong> this Section is appl<strong>ic</strong>able to compute immediate (time-<strong>in</strong>dependent)<br />

elasto-plast<strong>ic</strong> <strong>in</strong>stability of th<strong>in</strong> shells without stiffeners. It requires that an elast<strong>ic</strong> analysis (B<br />

3023) be per<strong>for</strong>med on the nom<strong>in</strong>al or "perfect" geometry of the structure subjected to the<br />

load<strong>in</strong>g <strong>in</strong> question (without load factors) and the follow<strong>in</strong>g quantities be determ<strong>in</strong>ed:<br />

The procedure consists of<br />

P m = primary membrane stress <strong>in</strong>tensity.<br />

PL + Pb<br />

= primary membrane plus bend<strong>in</strong>g stress <strong>in</strong>tensity.<br />

1) obta<strong>in</strong><strong>in</strong>g the elast<strong>ic</strong> bifurcation stress <strong>in</strong>tensities of the nom<strong>in</strong>al geometry us<strong>in</strong>g<br />

standard eigenvalue analysis,<br />

2) determ<strong>in</strong><strong>in</strong>g the <strong>in</strong>stability stress <strong>in</strong>tensities from the bifurcation stress <strong>in</strong>tensities,<br />

us<strong>in</strong>g factors that depend on the "imperfection" or deviation of the shell from its<br />

nom<strong>in</strong>al geometry and the yield stress of the material,<br />

3) check<strong>in</strong>g that the marg<strong>in</strong> between the <strong>in</strong>stability stress <strong>in</strong>tensities and the primary<br />

stress <strong>in</strong>tensities equal or exceed the load factors.<br />

1. Bifurcation analysis<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 52


ITER G 74 MA 8 01-05-28 W0.2<br />

The elast<strong>ic</strong> bifurcation stress <strong>in</strong>tensities are determ<strong>in</strong>ed (e.g., by f<strong>in</strong>ite-element analysis) by<br />

solv<strong>in</strong>g an eigenvalue problem, lead<strong>in</strong>g to a m<strong>in</strong>imum positive eigenvalue lc wh<strong>ic</strong>h is the<br />

multiplier by wh<strong>ic</strong>h the primary stress <strong>components</strong> need be multiplied to achieve bifurcation.<br />

If the eigenvalue analysis yields a negative value of lc, <strong>in</strong>stability cannot occur and further<br />

<strong>in</strong>stability analysis should be discont<strong>in</strong>ued. Otherwise, the follow<strong>in</strong>g bifircation stress<br />

<strong>in</strong>tensities are established:<br />

2. Instability stress <strong>in</strong>tensities<br />

Crit<strong>ic</strong>al membrane stress <strong>in</strong>tensity<br />

( Pm) c<br />

= lc1 Pm<br />

(1)<br />

Crit<strong>ic</strong>al membrane plus bend<strong>in</strong>g stress <strong>in</strong>tensity<br />

( L b ) =<br />

c c ( L + b )<br />

P + P l 2 P P<br />

(2)<br />

To determ<strong>in</strong>e the <strong>in</strong>stability stress <strong>in</strong>tensities, first the geometr<strong>ic</strong>al imperfection of the shell<br />

has to be characterized from the tolerances given on the draw<strong>in</strong>gs <strong>in</strong> the Component Data<br />

File.<br />

This imperfection may be def<strong>in</strong>ed as be<strong>in</strong>g the largest distance separat<strong>in</strong>g the two mean<br />

surfaces correspond<strong>in</strong>g to the nom<strong>in</strong>al geometry and a possible true geometry def<strong>in</strong>ed us<strong>in</strong>g<br />

the tolerances. This distance (d) may be evaluated on the segment perpend<strong>ic</strong>ular to the<br />

nom<strong>in</strong>al geometry (Figure B 3421-1).<br />

d<br />

M' M<br />

M = Nom<strong>in</strong>al geometry<br />

M' = Possible true geometry based on tolerances<br />

MM' is perpend<strong>ic</strong>ular to the mean surface of the nom<strong>in</strong>al geometry<br />

d = Max (MM')<br />

Figure B 3421-1: Distance ÒdÓ of imperfection <strong>in</strong> geometry<br />

To simplify the analysis, it is possible to consider only a fraction of the previous value if it<br />

can be shown that the neglected fraction of the tolerance has no effect on <strong>in</strong>stability. Thus,<br />

<strong>for</strong> a tube subjected to external pressure, only the ovalization tolerance is to be considered <strong>for</strong><br />

def<strong>in</strong><strong>in</strong>g the imperfection, ignor<strong>in</strong>g the tolerance on the mean diameter.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 53


ITER G 74 MA 8 01-05-28 W0.2<br />

The deviation <strong>in</strong>dex is def<strong>in</strong>ed <strong>in</strong> terms of the distance d and the shell th<strong>ic</strong>kness h by<br />

d = d/h<br />

The reduction factors to be applied to the bifurcation stress <strong>in</strong>tensities to account <strong>for</strong> the<br />

effects of deviation <strong>in</strong>dex and plast<strong>ic</strong>ity on the <strong>in</strong>stability stress <strong>in</strong>tensities are determ<strong>in</strong>ed<br />

from buckl<strong>in</strong>g diagrams (see Figures B 3421.1.2-1a, -1b, -2a, and -2b <strong>in</strong> the follow<strong>in</strong>g<br />

section) wh<strong>ic</strong>h depend on<br />

- whether the post-buckl<strong>in</strong>g behaviour of the structure is stable or unstable<br />

- material<br />

- temperature<br />

To use the buckl<strong>in</strong>g diagrams, values of the follow<strong>in</strong>g parameters are needed:<br />

- the deviation <strong>in</strong>dex d,<br />

( )<br />

- x = ( P ) S or x = P + P S<br />

m m c y L+ b L b c y<br />

Straight l<strong>in</strong>es Qm (or QL+b) with slopes xm (or xL+b) are drawn through the orig<strong>in</strong> of a<br />

buckl<strong>in</strong>g diagram, and their <strong>in</strong>tersections with the curves D def<strong>in</strong>ed by d = constant are noted,<br />

as shown <strong>in</strong> Figure B 3421-2 below.<br />

Figure B 3421-2: Use of buckl<strong>in</strong>g diagrams<br />

The <strong>in</strong>tersections have coord<strong>in</strong>ate values given by<br />

and<br />

( )<br />

X = ( s ) ( P ) or X = ( s ) P + P (3)<br />

m m I m c L+ b L+ b I L b c<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 54


ITER G 74 MA 8 01-05-28 W0.2<br />

Y s S or Y s S<br />

(4)<br />

= ( ) = ( )<br />

m m I y L+ b L+ b I y<br />

From Eqs. 1-4, the values of the <strong>in</strong>stability membrane ( sm ) and membrane plus bend<strong>in</strong>g<br />

I<br />

( ) stress <strong>in</strong>tensities and their correspond<strong>in</strong>g load factors can be evaluated as follows:<br />

sL+ b I<br />

and<br />

Load factors<br />

( l ) =<br />

m I<br />

( )<br />

s<br />

P<br />

m I<br />

m<br />

( s + )<br />

( lL+<br />

b) =<br />

I P + P<br />

L b I<br />

L b<br />

The f<strong>in</strong>al step <strong>in</strong> the buckl<strong>in</strong>g analysis consists of verify<strong>in</strong>g that the buckl<strong>in</strong>g load factors are<br />

equal to or greater than the <strong>design</strong> load factors (GL) as follows:<br />

and<br />

GL £ ( l m) (5)<br />

I<br />

GL L+ b I<br />

£ ( )<br />

l (6)<br />

The <strong>design</strong> load factors (GL) <strong>for</strong> the various serv<strong>ic</strong>e levels are tabulated <strong>in</strong> Table IC 3421-1.<br />

B 3421.1.2 Buckl<strong>in</strong>g diagrams<br />

The buckl<strong>in</strong>g diagrams <strong>for</strong> unstable post-buckl<strong>in</strong>g behaviour are given <strong>in</strong> Figures B 3421-3a<br />

and -3b, below. The buckl<strong>in</strong>g diagrams <strong>for</strong> stable post-buckl<strong>in</strong>g behaviour are given <strong>in</strong><br />

Figures B 3421-4a and -4b. The figures labeled "a" are the complete diagrams. The figures<br />

labeled "b" provide more detail near the orig<strong>in</strong>. These figures are appl<strong>ic</strong>able <strong>for</strong> annealed<br />

unirradiated type 316 sta<strong>in</strong>less steel between 20 - 700¡C. Note that be<strong>for</strong>e decid<strong>in</strong>g on wh<strong>ic</strong>h<br />

figure to use, the <strong>design</strong>er has to make a determ<strong>in</strong>ation as to whether the post-buckl<strong>in</strong>g<br />

equilibrium is stable or unstable. For example, the post-buckl<strong>in</strong>g behaviour is unstable <strong>for</strong><br />

circular cyl<strong>in</strong>dr<strong>ic</strong>al shells under axial compression or semi-spher<strong>ic</strong>al shell under external<br />

pressure. On the other hand, post-buckl<strong>in</strong>g behaviour is stable <strong>for</strong> circular cyl<strong>in</strong>dr<strong>ic</strong>al shells<br />

under external pressure or flat plate under <strong>in</strong>-plane compressive load<strong>in</strong>g. If a determ<strong>in</strong>ation<br />

cannot be made, to be conservative, the buckl<strong>in</strong>g curves <strong>for</strong> unstable post-buckl<strong>in</strong>g behaviour<br />

should be used.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 55


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure B 3421-3a: Unstable post-buckl<strong>in</strong>g behaviour diagram<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 56


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure B 3421-3b: Unstable post-buckl<strong>in</strong>g behaviour diagram, details<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 57


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure B 3421-4a: Stable post-buckl<strong>in</strong>g behaviour diagram<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 58


ITER G 74 MA 8 01-05-28 W0.2<br />

Figure B 3421-4b: Stable post-buckl<strong>in</strong>g behaviour diagram, details<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 59


ITER G 74 MA 8 01-05-28 W0.2<br />

B 3421.1.2 Stra<strong>in</strong>-controlled buckl<strong>in</strong>g limits<br />

Even though it is self-limit<strong>in</strong>g, stra<strong>in</strong>-controlled buckl<strong>in</strong>g must be avoided to guard aga<strong>in</strong>st<br />

failure by fatigue, excessive stra<strong>in</strong> (ratchet<strong>in</strong>g), and <strong>in</strong>teraction with load-controlled<br />

<strong>in</strong>stability. For purely stra<strong>in</strong>-controlled buckl<strong>in</strong>g, the effects of geometr<strong>ic</strong>al imperfections<br />

and tolerances, whether <strong>in</strong>itially present or <strong>in</strong>duced by serv<strong>ic</strong>e, need not be considered <strong>in</strong> the<br />

calculation of the <strong>in</strong>stability stra<strong>in</strong>. The procedure of B 3421.1.1 can be used to determ<strong>in</strong>e<br />

the bifurcation stra<strong>in</strong> and the <strong>in</strong>stability stra<strong>in</strong> after replac<strong>in</strong>g<br />

Pmand P L + Pbby QLand Q L + Qb,<br />

respectively, and the <strong>design</strong> load factors GL by GS <strong>in</strong><br />

Table IC 3422-1<br />

For thermally-<strong>in</strong>duced, stra<strong>in</strong>-controlled buckl<strong>in</strong>g, the stra<strong>in</strong> factor is applied to the loads<br />

<strong>in</strong>duced by thermal stra<strong>in</strong>s. To determ<strong>in</strong>e the buckl<strong>in</strong>g (bifurcation) stra<strong>in</strong>, it may be<br />

necessary to artif<strong>ic</strong>ially <strong>in</strong>duce high stra<strong>in</strong>s concurrent with the use of realist<strong>ic</strong> stiffness<br />

properties. The use of an "adjusted" thermal expansion coeff<strong>ic</strong>ient is one technique <strong>for</strong><br />

enhanc<strong>in</strong>g the applied stra<strong>in</strong>s without affect<strong>in</strong>g the associated stiffness properties. The<br />

plast<strong>ic</strong>ity correction <strong>for</strong> <strong>in</strong>stability stress <strong>in</strong>tensities may be obta<strong>in</strong>ed from the curves <strong>in</strong> the<br />

buckl<strong>in</strong>g diagrams (Figs B 3421-3a/4b that correspond to d=0).<br />

Although stra<strong>in</strong> factors <strong>for</strong> stra<strong>in</strong>-controlled buckl<strong>in</strong>g are less than load factors <strong>for</strong> loadcontrolled<br />

buckl<strong>in</strong>g, <strong>for</strong> conditions where signif<strong>ic</strong>ant elast<strong>ic</strong> follow-up may occur or where<br />

load-controlled and stra<strong>in</strong>-controlled buckl<strong>in</strong>g may <strong>in</strong>teract, the load factors appl<strong>ic</strong>able to<br />

load-controlled buckl<strong>in</strong>g shall be applied.<br />

B 3421.2 Elasto-plast<strong>ic</strong> analysis (Immediate elasto-plast<strong>ic</strong> <strong>in</strong>stability)<br />

If the elast<strong>ic</strong> buckl<strong>in</strong>g limits cannot be met, elasto-plast<strong>ic</strong> buckl<strong>in</strong>g analysis may be<br />

conducted. By its very nature it is a much more complex set of calculations than elast<strong>ic</strong><br />

analysis. The bas<strong>ic</strong> procedure is as follows.<br />

1) Obta<strong>in</strong> the elast<strong>ic</strong> bifurcation stress <strong>in</strong>tensities and modes of the nom<strong>in</strong>al geometry<br />

us<strong>in</strong>g standard eigenvalue analysis.<br />

2) Determ<strong>in</strong>e a modified geometry by <strong>in</strong>corporat<strong>in</strong>g a defect with a shape of the<br />

bifurcation mode correspond<strong>in</strong>g to the lowest bifurcation load and with a<br />

maximum amplitude (Fig. B 3400) that is consistent with the tolerances given <strong>in</strong><br />

the Component Data File.<br />

3) If the maximum stress correspond<strong>in</strong>g to the elast<strong>ic</strong> bifurcation load<strong>in</strong>g is <strong>in</strong> the<br />

plast<strong>ic</strong> range, the modified geometry may have to be constructed us<strong>in</strong>g an elastoplast<strong>ic</strong><br />

bifurcation mode, wh<strong>ic</strong>h should be determ<strong>in</strong>ed us<strong>in</strong>g the follow<strong>in</strong>g<br />

procedure.<br />

- The pre-buckl<strong>in</strong>g equilibrium states of the structure subjected to proportional<br />

load<strong>in</strong>gs (lL) marked by a load<strong>in</strong>g parameter l are calculated. This<br />

calculation, made on the nom<strong>in</strong>al geometry, should take <strong>in</strong>to account the<br />

effects of geometr<strong>ic</strong>al non-l<strong>in</strong>earities and plast<strong>ic</strong>ity.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 60


ITER G 74 MA 8 01-05-28 W0.2<br />

- For each load<strong>in</strong>g value l, the coeff<strong>ic</strong>ient k by wh<strong>ic</strong>h the correspond<strong>in</strong>g stress<br />

state should be multiplied, <strong>in</strong> order to obta<strong>in</strong> bifurcation, must be calculated.<br />

If the stress state is "plast<strong>ic</strong>", this coeff<strong>ic</strong>ient should be calculated by replac<strong>in</strong>g<br />

Young's modulus by a modified modulus, at each po<strong>in</strong>t wh<strong>ic</strong>h has become<br />

"plast<strong>ic</strong>". It is recommended that the tangent modulus at the po<strong>in</strong>t concerned<br />

be adopted as the modified modulus. Other values may also be used provided<br />

that the cho<strong>ic</strong>e can be justified.<br />

- The elasto-plast<strong>ic</strong> bifurcation load (lcL) is the load at wh<strong>ic</strong>h the multipl<strong>ic</strong>ation<br />

coeff<strong>ic</strong>ient k is equal to 1; the de<strong>for</strong>mation pattern (eigenmode) of this mode is<br />

the bifurcation load to be taken <strong>in</strong>to consideration.<br />

- If the lowest elast<strong>ic</strong> bifurcation load (Cbe) is much higher than the elastoplast<strong>ic</strong><br />

bifurcation load (Cbi), a new "modified" geometry shall be constructed<br />

like the one <strong>for</strong> the elast<strong>ic</strong> case, but us<strong>in</strong>g the elasto-plast<strong>ic</strong> bifurcation mode.<br />

Unless otherwise justified, the elasto-plast<strong>ic</strong> bifurcation mode must be used if<br />

the elast<strong>ic</strong> bifurcation load is 10 times greater than the elasto-plast<strong>ic</strong><br />

bifurcation load (Cbe > 10 Cbi).<br />

4) Conduct an <strong>in</strong>cremental large-displacement elasto-plast<strong>ic</strong> nonl<strong>in</strong>ear analysis on<br />

the modified geometry us<strong>in</strong>g the m<strong>in</strong>imum stress-stra<strong>in</strong> curves of the material and<br />

loads applied proportionally after <strong>in</strong>corporat<strong>in</strong>g the appropriate load and stra<strong>in</strong><br />

factors.<br />

Instability is deemed to have occurred if the load-displacement curve reaches a maximum or<br />

the maximum displacement exceeds acceptable limit, be<strong>for</strong>e the f<strong>in</strong>al load is reached.<br />

Note: If f<strong>in</strong>ite element analysis is used, the bas<strong>ic</strong> approach is to <strong>in</strong>crement the applied loads<br />

until the solution beg<strong>in</strong>s to diverge. Care must be taken to use a suff<strong>ic</strong>iently f<strong>in</strong>e load<br />

<strong>in</strong>crement as the load approaches the expected buckl<strong>in</strong>g load. If the load <strong>in</strong>crement is too<br />

coarse the buckl<strong>in</strong>g load may not be accurate. It is important to keep <strong>in</strong> m<strong>in</strong>d that an<br />

unconverged solution does not necessarily mean that the load has reached a maximum. It<br />

could also be caused by numer<strong>ic</strong>al <strong>in</strong>stability. Track<strong>in</strong>g the load-displacement history of the<br />

component can be helpful <strong>in</strong> decid<strong>in</strong>g whether an unconverged load step represents actual<br />

buckl<strong>in</strong>g or whether it reflects some other problem.<br />

B 3500 High Temperature Rules<br />

(Will be issued at a later date.)<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 61


ITER G 74 MA 8 01-05-28 W0.2<br />

BÊ3800 DESIGN RULES FOR BOLTED JOINTS<br />

BÊ3810 Methods of analysis<br />

BÊ3811 Elast<strong>ic</strong> analysis<br />

Detailed stress analysis of a bolted jo<strong>in</strong>t can be quite complex depend<strong>in</strong>g on the level of<br />

detail to wh<strong>ic</strong>h the geometry and the load<strong>in</strong>g of the jo<strong>in</strong>t is modelled. The complexity arises<br />

from the fact that it is a non-<strong>in</strong>tegral connection us<strong>in</strong>g <strong>in</strong>itially prestressed bolts wh<strong>ic</strong>h<br />

requires that a discont<strong>in</strong>uity stress analysis be conducted to satisfy the compatibility of<br />

de<strong>for</strong>mations with<strong>in</strong> the jo<strong>in</strong>t. This art<strong>ic</strong>le provides guidel<strong>in</strong>es <strong>for</strong> a general procedure <strong>for</strong><br />

analyz<strong>in</strong>g a bolted jo<strong>in</strong>t. The <strong>design</strong>er may choose any other procedure that can be shown to<br />

give conservative results.<br />

In order to per<strong>for</strong>m a discont<strong>in</strong>uity analysis of a bolted jo<strong>in</strong>t, the follow<strong>in</strong>g <strong>in</strong><strong>for</strong>mation must<br />

be available:<br />

(1) the dimensions of the jo<strong>in</strong>t (<strong>in</strong>clud<strong>in</strong>g supports) and bolts,<br />

(2) the material properties such as Young's Moduli, Poisson's ratios, and thermal<br />

expansion coeff<strong>ic</strong>ients of all <strong>components</strong>,<br />

(3) mechan<strong>ic</strong>al loads, such as pressure, dead weight, bolt loads, and other externally<br />

applied <strong>for</strong>ces and moments on the jo<strong>in</strong>t, and<br />

(4) temperature distribution <strong>in</strong> the component parts.<br />

The discont<strong>in</strong>uity analysis of a bolted jo<strong>in</strong>t can be per<strong>for</strong>med us<strong>in</strong>g standard procedures <strong>for</strong><br />

analysis of stat<strong>ic</strong>ally <strong>in</strong>determ<strong>in</strong>ate structures satisfy<strong>in</strong>g equilibrium of <strong>for</strong>ces and<br />

compatibility of displacements at each <strong>in</strong>terface. However, <strong>for</strong> a complex structure with<br />

many bolts, <strong>in</strong>terfaces, and shear keys, such an analysis may sometimes be unwieldy. In such<br />

cases, a pract<strong>ic</strong>al recourse is to analyze the structure by f<strong>in</strong>ite element analysis. Although<br />

many commercial f<strong>in</strong>ite element codes provide sophist<strong>ic</strong>ated gap elements to simulate<br />

realist<strong>ic</strong> <strong>in</strong>terfaces with fr<strong>ic</strong>tion, such analyses are <strong>in</strong>herently nonl<strong>in</strong>ear and may be prone to<br />

convergence problems. S<strong>in</strong>ce most jo<strong>in</strong>ts are <strong>design</strong>ed with bolt preloads such that gaps do<br />

not develop at the <strong>in</strong>terfaces due to serv<strong>ic</strong>e load<strong>in</strong>g, simpler f<strong>in</strong>ite element models should be<br />

adequate.<br />

Depend<strong>in</strong>g on the f<strong>in</strong>ite element model of the jo<strong>in</strong>t, either beam elements or solid elements<br />

may be used to model the bolts at the m<strong>in</strong>imum bolt diameter or at the root of threads. The<br />

detailed geometry of the threads <strong>in</strong> either the bolts or the flange need not be <strong>in</strong>cluded <strong>in</strong> the<br />

model. Designers should use judgment to set the appropriate boundary conditions at the jo<strong>in</strong>t<br />

<strong>in</strong>terfaces as well as the <strong>in</strong>terfaces between the flange and the bolts. In all cases, the normal<br />

displacements at the <strong>in</strong>terfaces should be made cont<strong>in</strong>uous and, depend<strong>in</strong>g on the condition<br />

of the <strong>in</strong>terfaces, either the tangential displacements should also be made cont<strong>in</strong>uous (no<br />

<strong>in</strong>terfacial slippage) or the tangential shear stresses should be set equal to zero (lubr<strong>ic</strong>ated<br />

<strong>in</strong>terface).<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 62


ITER G 74 MA 8 01-05-28 W0.2<br />

The <strong>in</strong>itial stresses due to bolt preload can be obta<strong>in</strong>ed by f<strong>in</strong>ite element analysis us<strong>in</strong>g a<br />

f<strong>ic</strong>titious temperature drop of the bolts to simulate a lack of fit. The validity of the results<br />

should be checked by ensur<strong>in</strong>g that the normal stresses are compressive at each <strong>in</strong>terface.<br />

A major unknown <strong>in</strong> a bolted jo<strong>in</strong>t analysis is the redistribution of the normal pressure and<br />

tangential stress at the jo<strong>in</strong>t <strong>in</strong>terfaces due to serv<strong>ic</strong>e load<strong>in</strong>g. As mentioned earlier, most<br />

jo<strong>in</strong>ts are <strong>design</strong>ed such that gaps do not develop at the <strong>in</strong>terfaces due to serv<strong>ic</strong>e load<strong>in</strong>g.<br />

Thus, as a first step, a global stress analysis of the jo<strong>in</strong>t should be conducted under all applied<br />

mechan<strong>ic</strong>al and thermal load<strong>in</strong>g (without the bolt pre-load<strong>in</strong>g) and us<strong>in</strong>g either the no<br />

slippage <strong>in</strong>terfaces or lubr<strong>ic</strong>ated <strong>in</strong>terfaces, as discussed earlier. By superposition pr<strong>in</strong>ciple,<br />

the total stresses <strong>in</strong> the structure are obta<strong>in</strong>ed by add<strong>in</strong>g the stresses due to serv<strong>ic</strong>e load<strong>in</strong>g to<br />

those due to bolt preloads. The <strong>design</strong> bolt preloads should be chosen such that the total<br />

normal stresses rema<strong>in</strong> compressive at all <strong>in</strong>terfaces under all load<strong>in</strong>gs. Satisfy<strong>in</strong>g this<br />

condition also ensures that the simplified analysis is valid. If gaps over signif<strong>ic</strong>ant areas of<br />

the <strong>in</strong>terfaces cannot be avoided due to some load<strong>in</strong>gs, nonl<strong>in</strong>ear gap elements may be<br />

considered at the <strong>in</strong>terfaces.<br />

Often the stiffness of the flange is much greater than that of the bolts. In such cases, applied<br />

mechan<strong>ic</strong>al loads (such as plasma disruption-<strong>in</strong>duced electromagnet<strong>ic</strong> loads) can be<br />

accommodated <strong>in</strong> the jo<strong>in</strong>t by a reduction of <strong>in</strong>terfacial pressure with relatively little change<br />

<strong>in</strong> the bolt stress. However, bolt stresses can <strong>in</strong>crease signif<strong>ic</strong>antly if gaps develop at the<br />

<strong>in</strong>terfaces. On the other hand, bolt stresses can vary signif<strong>ic</strong>antly with differential thermal<br />

expansion (either due to temperature difference and/or due to difference <strong>in</strong> thermal expansion<br />

coeff<strong>ic</strong>ient) between the flange and the bolts even <strong>in</strong> the absence of gaps. Thus from a<br />

standpo<strong>in</strong>t of fatigue of the bolts <strong>in</strong> ITER blanket modules, cycl<strong>ic</strong> thermal stresses of the<br />

bolts are likely to be more crit<strong>ic</strong>al than cycl<strong>ic</strong> stresses due to electromagnet<strong>ic</strong> load<strong>in</strong>g dur<strong>in</strong>g<br />

plasma disruptions.<br />

In fatigue analysis of the bolts, the range of the maximum concentrated stress at the root of<br />

the threads is needed. This can be obta<strong>in</strong>ed by a local detailed elast<strong>ic</strong> f<strong>in</strong>ite element analysis<br />

of the bolt (<strong>in</strong>clud<strong>in</strong>g the threads) us<strong>in</strong>g the results from the global analysis to set the<br />

boundary conditions. Alternatively, a fatigue strength reduction factor (Kf) (IC 2753) can be<br />

used to estimate the maximum stress. In both cases, Neuber analysis (B 3024.1.3) can be<br />

used to estimate the <strong>in</strong>fluence of plast<strong>ic</strong> flow on the peak stress and stra<strong>in</strong> ranges, as<br />

discussed <strong>in</strong> IC 3851.2.1. Because of the presence of a high tensile mean stress, wh<strong>ic</strong>h is<br />

known to have an adverse effect on fatigue life, a mean stress correction based on the<br />

Goodman equation has been recommended.<br />

B 3812 Simplified elast<strong>ic</strong> analysis<br />

A bolted connection can be visualized as two spr<strong>in</strong>gs <strong>in</strong> parallel, one with stiffness K of the<br />

assembled parts and the other with stiffness KB of the bolts. Expressions <strong>for</strong> stiffness factors<br />

based on simplified uniaxial analysis of various types of jo<strong>in</strong>ts are given <strong>in</strong> Appendix A6 of<br />

RCC-MR. Alternatively, f<strong>in</strong>ite element analysis may be used to derive more accurate<br />

stiffness factors <strong>for</strong> the connection. F<strong>in</strong>ite element analysis will also be needed <strong>for</strong><br />

assemblies <strong>in</strong> wh<strong>ic</strong>h the assembled parts and bolts experience signif<strong>ic</strong>ant bend<strong>in</strong>g due to<br />

applied load<strong>in</strong>g.<br />

If bend<strong>in</strong>g of the assembled parts and the bolts can be neglected, a uniaxial analysis can be<br />

used. This simplified model may be appl<strong>ic</strong>able to a number of bolted assemblies <strong>for</strong> wh<strong>ic</strong>h<br />

precise calculations are not needed.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 63


ITER G 74 MA 8 01-05-28 W0.2<br />

B 3812.1 Simplified uniaxial analysis<br />

Consider the simplified bolted assembly (with two parts connected by a threaded screw)<br />

shown <strong>in</strong> Fig. B 3812-1. Although only a s<strong>in</strong>gle screw connect<strong>in</strong>g two parts are shown, the<br />

follow<strong>in</strong>g analysis can be extended to jo<strong>in</strong>ts with n ident<strong>ic</strong>al screws <strong>in</strong> parallel connect<strong>in</strong>g J<br />

number of parts <strong>in</strong> series. The stiffnesses of the <strong>in</strong>dividual bolts and the parts are given by<br />

K<br />

B0<br />

EB0AB = , K<br />

L<br />

B<br />

i0<br />

Ei0Ai = , i = 1, 2,..., J<br />

L<br />

i<br />

where EB0, and Ei0 are the moduli of elast<strong>ic</strong>ity of the bolt, and part i, respectively at <strong>in</strong>itial<br />

temperature q0. Denot<strong>in</strong>g the <strong>in</strong>itial comb<strong>in</strong>ed stiffness of the parts by K0,<br />

i= J<br />

1 1<br />

= å K Ki 0 i = 1 0<br />

Denot<strong>in</strong>g the <strong>in</strong>itial elongation of the bolts by uB0, the total <strong>in</strong>itial compression of the parts<br />

by u0, and the <strong>in</strong>itial bolt preload (per bolt) by PB0,<br />

u<br />

u<br />

B0<br />

0<br />

P<br />

=<br />

K<br />

nP<br />

=<br />

K<br />

B0<br />

B0<br />

B0<br />

B0<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 64


ITER G 74 MA 8 01-05-28 W0.2<br />

A<br />

A<br />

A<br />

p I<br />

1<br />

B<br />

2<br />

P B<br />

Fig. B 3812-1 Simplified bolted assembly used <strong>for</strong> uniaxial analysis.<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 65<br />

L 1<br />

L 2<br />

B 3812.1.1 Effects of a temperature rise<br />

The free elongations of the bolts and the parts due to temperature rises DTB and<br />

DT, respectively, from the <strong>in</strong>itial temperature T0 are given by<br />

DuB = LBaB DTB<br />

Dui = Li ai DTi<br />

<strong>for</strong> the bolts and<br />

L B<br />

P B<br />

<strong>for</strong> the ith part, i = 1, 2, ..., J


ITER G 74 MA 8 01-05-28 W0.2<br />

where aB and ai are the thermal expansion coeff<strong>ic</strong>ients of the bolts and part<br />

i at temperatures T0+DTB andT0+DTi, respectively.<br />

The change <strong>in</strong> the bolt preload DPB (per bolt) due to the thermal expansion mismatch is given<br />

by<br />

KK éi=<br />

J<br />

ù<br />

B<br />

DPB= êå<br />

LiaiDTi - LBaBDTBú K + nKBëêi=<br />

1<br />

ûú<br />

where the stiffnesses are evaluated with Young's moduli at the f<strong>in</strong>al<br />

temperatures and a positive sign denotes <strong>in</strong>crease <strong>in</strong> tension and a negative<br />

sign denotes decrease <strong>in</strong> tension.<br />

If we <strong>in</strong>clude the change <strong>in</strong> the <strong>in</strong>itial preload due to change <strong>in</strong> Young's moduli at f<strong>in</strong>al<br />

temperatures, the f<strong>in</strong>al bolt preload (per bolt) is given by<br />

KKB<br />

K + nK<br />

PB = DPB<br />

+ ×<br />

KK K + nK<br />

0 B0<br />

0 B0<br />

B<br />

B 3812.1.2 Effects of an external load applied to the bolted<br />

assembly<br />

If a tensile (or compressive) load N is applied along the symmetry axis of the bolted jo<strong>in</strong>t, the<br />

load <strong>in</strong> the bolts <strong>in</strong>creases (or decreases) by DNB (per bolt) and that <strong>in</strong> the assembled parts<br />

decreases (or <strong>in</strong>creases) by DN so that<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 66<br />

× P<br />

- the total tensile load <strong>in</strong> each bolt = PB + DNB<br />

- the total compressive load <strong>in</strong> the parts assembled = nPB - DN<br />

- equilibrium of <strong>for</strong>ces gives N = nDNB + DN<br />

- compatibility of <strong>in</strong>cremental axial de<strong>for</strong>mations of the bolts and parts give<br />

DNBDN =<br />

K K<br />

B<br />

- total axial load <strong>in</strong> each bolt = P<br />

B<br />

B0<br />

K<br />

K nK N B<br />

+<br />

+<br />

- total compressive load <strong>in</strong> assembled parts = nP<br />

B<br />

B<br />

K<br />

K nK N<br />

-<br />

+<br />

nK<br />

- to ensure that the assembled parts rema<strong>in</strong> <strong>in</strong> compression, N < nP<br />

æ<br />

B 1 +<br />

è K<br />

Note: Validity of the above simplified analyses must be checked by ensur<strong>in</strong>g that<br />

(1) the applied mechan<strong>ic</strong>al and thermal loads do not cause the jo<strong>in</strong>t to become loose, i.e.,<br />

jo<strong>in</strong>t contact pressure always rema<strong>in</strong>s positive<br />

B<br />

B<br />

ö<br />

ø


ITER G 74 MA 8 01-05-28 W0.2<br />

(2) bend<strong>in</strong>g of the assembled parts do not cause a pry<strong>in</strong>g action on the bolts wh<strong>ic</strong>h may<br />

lead to bend<strong>in</strong>g and <strong>in</strong>creased tensile stress of the bolts. The flange th<strong>ic</strong>knesses, bolt<br />

spac<strong>in</strong>g, distance from free edge and preload should be chosen so that such pry<strong>in</strong>g<br />

action is reduced to a m<strong>in</strong>imum.<br />

B 3812.1.3 Effects of an external moment applied to the bolted<br />

assembly<br />

Consider a bolted assembly subjected to a bend<strong>in</strong>g moment My as shown <strong>in</strong> Fig. B 3812-2.<br />

Assume that the section rema<strong>in</strong>s flat but rotates by an angle qy and that the <strong>in</strong>terfaces<br />

transmit the moment by a readjustment of the bolt loads and a redistribution of the <strong>in</strong>terfacial<br />

pressures wh<strong>ic</strong>h rema<strong>in</strong> positive at all locations of the <strong>in</strong>terfaces. Also, because of the<br />

rotation of the section, each bolt will be subjected to a bend<strong>in</strong>g moment mBy wh<strong>ic</strong>h is applied<br />

to the bolt by a redistribution of the <strong>in</strong>terfacial pressure under the bolt head (or nut). Us<strong>in</strong>g a<br />

similar uniaxial model as be<strong>for</strong>e, it can be shown that<br />

q y<br />

My<br />

=<br />

K 2 2<br />

nR + K åx + x dA - x nKB + K<br />

A ò<br />

B<br />

i= n<br />

B<br />

i = 1<br />

2<br />

Bi<br />

A<br />

( )<br />

where A is the section area and RB is the rotational stiffness of the bolts. If<br />

the section has a symmetry about the Y axis and the orig<strong>in</strong> of the x axis lies<br />

on it, then x =0, if not, then it is given by<br />

i= n<br />

K<br />

K x<br />

A<br />

x<br />

xdA<br />

B å Bi + ò<br />

i = 1 A<br />

=<br />

.<br />

nK + K<br />

B<br />

In above K 1<br />

= .<br />

i= J<br />

A Li<br />

å E<br />

i = 1<br />

i<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 67


ITER G 74 MA 8 01-05-28 W0.2<br />

Y<br />

M<br />

y<br />

x<br />

Bi<br />

ith Bolt<br />

Fig. B 3812-2 Bolted assembly subjected to a bend<strong>in</strong>g moment<br />

The changes <strong>in</strong> the bolt axial load and the maximum reduction <strong>in</strong> the <strong>in</strong>terfacial pressure are<br />

given by<br />

DP = K q x -x<br />

Dp<br />

Bi B y Bi<br />

( )<br />

Kqy<br />

= - x -x<br />

A<br />

( )<br />

max max<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 68<br />

q y<br />

y<br />

Bi<br />

where xmax is the maximum x value <strong>for</strong> the section.<br />

The bend<strong>in</strong>g moment on each bolt is given by<br />

mBy = RBqy,<br />

where RB is the rotational stiffness of the bolt.<br />

For comput<strong>in</strong>g RB, the bend<strong>in</strong>g of the bolt head (and the washer, if any) can be neglected,<br />

i.e.,<br />

R<br />

B<br />

EBd =<br />

L<br />

p<br />

32<br />

B<br />

4<br />

X


ITER G 74 MA 8 01-05-28 W0.2<br />

where d is the nom<strong>in</strong>al diameter of the bolt.<br />

Similar expressions can be derived <strong>for</strong> an applied moment Mx about the x axis. To m<strong>in</strong>imize<br />

the stresses due to the applied bend<strong>in</strong>g moment, the bolts should be arranged as<br />

symmetr<strong>ic</strong>ally as possible.<br />

For the above analysis to rema<strong>in</strong> valid, the total <strong>in</strong>terfacial pressure must not be negative<br />

anywhere <strong>in</strong> the section or under the bolt heads. The m<strong>in</strong>imum total pressure at the <strong>in</strong>terfaces<br />

between the parts assembled is (<strong>in</strong>clud<strong>in</strong>g the contribution of the applied axial <strong>for</strong>ce N)<br />

p<br />

( ) - -<br />

nPB - N<br />

A<br />

Kq<br />

A<br />

( x x)<br />

y<br />

m<strong>in</strong> =<br />

max<br />

Similarly, the m<strong>in</strong>imum total pressure under the bolt head can be estimated by assum<strong>in</strong>g a<br />

l<strong>in</strong>ear pressure distribution to give<br />

p<br />

BH<br />

( )<br />

P N<br />

B n m d<br />

m<strong>in</strong><br />

d d d d<br />

=<br />

4 + 32<br />

2 2 4<br />

p -<br />

p -<br />

( B ) -<br />

1<br />

By B<br />

B 1 4<br />

( )<br />

where dB is the <strong>in</strong>scribed bolt head diameter and d1 is the diameter of the<br />

hole.<br />

To ma<strong>in</strong>ta<strong>in</strong> positive pressures between the parts assembled and between the bolt head and<br />

the part, we must have<br />

pm<strong>in</strong> > 0<br />

and<br />

p BHm<strong>in</strong> > 0<br />

BÊ3813 Elasto-plast<strong>ic</strong> analysis<br />

Detailed elasto-plast<strong>ic</strong> analysis of a bolted jo<strong>in</strong>t can be highly complex and should be<br />

attempted only if the elast<strong>ic</strong> analysis rules cannot be satisfied. S<strong>in</strong>ce the superposition<br />

pr<strong>in</strong>ciple cannot be used <strong>in</strong> elasto-plast<strong>ic</strong> analysis, stresses due to bolt preloads as well as<br />

serv<strong>ic</strong>e loads have to be determ<strong>in</strong>ed together us<strong>in</strong>g an <strong>in</strong>cremental approach (B 3024).<br />

SDC-IC, Appendix B. Guidel<strong>in</strong>es <strong>for</strong> Analysis page 69

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!