CONTENTS Report No. 1: Evaluation of Design Margins for ASME Code Section VIII, Division 1 Executive Summary ...................................................... 1 A. Introduction ........................................................... 1 B. Service Experience .................................................... 2 C. Failure Modes Considered ............................................. 4 1. Ductile Rupture ...................................................... 4 2. Brittle Fracture ...................................................... 4 3. Fatigue ............................................................... 5 4. Incremental Collapse ................................................. 5 5. Elastic Instability .................................................... 5 6. Plastic Instability ...............•.................................... 5 7. Creep Rupture ........................................•.............. 6 8. Creep Buckling ....................................................... 6 D. Improvements to the Code Since 1940's 1. Materials ............................................................. 6 2. Design ................................................................ 9 3. Fabrication ......................................................... 10 4. Welding ............................................................. 11 5.PWHT ............................................................... 11 6.NDE ................................................................. 11 7. Brittle Fracture .............•....................................... 12 E. Review of PVRC Burst Tests .......................................... 13 F. Review of Significant Code Paragraphs ............................... 14 G. Conclusions........................................................... 17 H. Suggestions for Revisions to Code Requirements ..................... 17 I. Recommendations ..................................................... 18 Bibliography . . . . . . . . . . . . . . . . . . . . ...................................... 18 ISBN No. 1-58145-442-2 Library of Congress Catalog Card Number: 99-64559 Copyright© 1999 by Welding Research Council, Inc. All Rights Reserved Printed in U.S.A. ii Report No. 2: Evaluation of Design Margins for ASME Code Section VIII, Divisions 1 and 2-Phase 2 Studies A. Introduction .......................................................... 20 B. Executive Summary .................................................. 20 C. Design Margins for Section VIII, Division 1 (Phase 1 Follow-up Studies) ............................................................. 25 1. Design Criteria of Section VIII, Division 1 ........................... 25 2. Effect of Design Margins on Fatigue Strength ....................... 26 3. Review of Code Case 2168 and Nozzle Reinfo~ement Requirements ....................................................... 27 4. The Effect of Lower Design Margin on Brittle Fracture of Part UHT Materials ............................................................ 28 D. Study of Vessel Test Results ........................................... 29 1. Purpose ............................................................. 29 2. CBI Reduced Girth Tests ............................................ 29 3. CBI Ellipsoidal Head Tests .......................................... 30 4. Containment Vessel Model Tests ..................................... 30 5. Praxair Torispherical Head Test..................................... 30 6. MPC Service Exposed Test Vessels With Local Thin Areas ........... 30 7. Conclusions ......................................................... 31 iii E. Evaluation of Design Margins for ASME Section VIII, Division 2 ...... 31 1. Review of the Failure Modes Considered in Section VIII, Division 2 .. 31 a. Ductile Rupture .............. ··· .. ································ 31 b. Brittle Fracture .................... · . · .. · · · · · · · · · · · · · · · · · · · · · · · · · · 32 c. Fatigue ..................... · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 32 d. Incremental Collapse ..................... · · .. · · · · · · · · · · · · · · · · · · · · 33 e. Elastic Instability .................. · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 33 f. Plastic Instability..................... · ... ························· 33 g. Excessive Deformation and Leakage .. : ........ : .. · · · · · · · · · · · · · · · · 33 2. Assessment of Significant Factors on Design Margms ............... 33 a. Design Details ........................ · .. · · · · · · · · · · · · · · · · · · · · · · · · · · 33 b. Ductile Rupture ................................... · .. · · · · · · · · · · · · · 33 c. Brittle Fracture ................................. · · · · · · · · · · · · · · · · · · 35 d. Yield Strength/UTS Ratios and Ductility ...................... · ... 40 e. Fabrication/NDE .............................. · · · · · · · · · · · · · · · · · · · · 44 3. Review of Vessel Failures ................................. ··.········ 45 4. Summary and Conclusions .............................. · · · · · · · · · · · · 45 F. Comparison ofASME Section VIII, Divisions 1and2 With Other Codes ....................................... ························· 46 1. Design Margins ................................. · · · · · · · · · · · · · · · · · · · · 46 a. General ............................. · · .. · · · · · · · · · · · · · · · · · · · · · · · · · · 46 b. ASME, Section VIII, Division 1 .................................... 48 c.ASME, Section VIII, Division 2 ......................... · · ·. · · · · · · · · 48 d. BS 5549 ................................... · · · · · · · · · · · · · · · · · · · · · · · · · 49 e. AD-Merkblatt................................... · · · · · · · · · · · · · · · · · · · 49 f. Stoomwezen ................................ · .. · · · · · · · · · · · · · · · · · · · · 49 g. CEN .......................................... ·· ········ · · · · ······ · 50 2. Design Rules ............................... · · · . · · · · · · · · · · · · · · · · · · · · · 50 a. Design for Internal Pressure ...................... · ... · · · · · · · · · · · · 50 b. Design of Dished Heads ........................... · . · · · · · · · · · · · · · · 51 c. Design of Flat Covers ........................... ··················· 51 d. Design of Flanges ......................... · · .. · · · · · · · · · · · · · · · · · · · · 51 e. Design of Openings ............................... · · . · · · · · · · · · · · · · · 51 f. Fatigue Analysis ............................ · .. · · · · · · · · · · · · · · · · · · · · 52 g. Design for External Pressure ...................................... 52 3. Design Details ..................................... · · · · · · · · · · · · · · · · · · 52 4. Special Service Considerations ...................................... 54 5. Materials Requirements ...................... · . · · · · · · · · · · · · · · · · · · · · · 55 6. Toughness Requirements ......................... · · · · · · · · · · · · · · · · · · · 57 7. Fabrication Requirements ......................... ·. · · · · · · · · · · · · · · · 61 a. Welded Joints .................................... · .... · · · · · · · · · · · · 61 b. Impact Testing of Welded Joints ................................... 62 c. Inspection of Welded Joints ............................. · .. • · · ·. · · 64 d Tolerances ..................... · ... · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 68 e:PWHT ............................................................. 69 f. Third Party Inspection ............................. · · · · · · · · · · · · · · · 71 8. Overpressure Testing Requirements ...........•.................... 72 9. Quality Control System .......•.............. · ... · · · · · · · · · · · · · · · · · · · 74 10. Summary and Conclusions .............................. ·· .. ······· 74 G. Recommendations for ASME Section VIII, Division 1 and Division 2 Revisions ........................ • · .. · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 79 H. Recommendations for Additional Studies ................ · · · · · · · · · · · · · 83 Bibliography ........................ · · .. · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 84 i"(T Report No. 1: Evaluation Design Margins Code Section VIII, Division 1 ASME E. Upitis and K. Mokhtarian Executive Summary Improvements to ASME Section VIII, Division 1 Code rules over the last 40-50 years, successful experience with vessels designed to the Code rules, improved materials and fabrication practices, more sophisticated design rules and toughness requirements have prompted a further review of the Section VIII, Division 1 requirements to see if there is a justification for a reduction in the present Code design margins from a factor of 4 to 3.5 on the ultimate tensile strength at temperatures below the creep range. This report discusses the successful experience with vessels designed to Section VIII, Division 1 rules, the failure modes considered in the Code rules and margins on design, improvements in the Code rules since the last reduction of the design margin, the Code paragraphs affected by changes in the design margins, areas which need a further review, and recommended additional work as a follow-up to this study to further justify the lower design margin. The Code rules apply to new construction only and have no control over the operation of vessels once they hit\re been placed in service. Failures have occurred in service from various causes and during overload testing, mainly in materials with poor notch toughness or because of poor details or fabrication practices. Failures during overload testing have now been largely eliminated because of the improved Code toughness rules and the recommendation that test water temperature be at least 30°F above the minimum design metal temperature. The main issues affecting the safety of new vessels are ductile rupture and brittle fracture. The work by B. F. Langer and burst tests at University of Kansas in the 1970's on test vessels with different strain hardening exponents have demonstrated that ductile rupture is highly unlikely with the reduced design margin of 3.5 on ultimate tensile strength. One of the most significant changes to the Code rules in Section VIII, Division 1 was the new toughness rules introduced in 1987. These are based on fracture mechanics considerations and past successful practices with Code vessels. Materials and welded joints are now subject to impact testing, unless exempted by the impact test exemption curves in Fig. UCS-66 and other Code rules. Although it is not possible to establish a single margin of safety against brittle fracture, it is unlikely that the safety of Section VIII, Division 1 vessels would be jeopardized by lowering the design margin from 4 to 3.5 on tensile strength, because of several factors: 1. The new Division 1 toughness rules. 2. Enhanced notch toughness of vessels resulting from improved steel-making practices and vessel manufacturing practices. 3. The successful experience with Division 2 vessels, designed to higher stresses, but subject to similar toughness requirements. 4. Improved design rules. A large part of the report has been devoted to the numerous improvements in the Code rules since the 1940's when the design margin was reduced from 5 to 4. Significant improvements have been made to materials, the Code design rules and design details, fabrication requirements, welding, postweld heat treat requirements, NDE requirements, and the Code toughness rules. New rules have been introduced where none existed in the early editions of the Code. These improvements have led to general improvement to the safety and reliability of Section VIII, Division 1 vessels, but no efforts have been made so far to try to justify a further reduction to the Code design margins, taking credit for the revisions and improvements to the Code rules. This report includes recommendations for more in-depth investigation in certain areas to provide further justification for reduction of the present design margins. Section H gives suggestions to the Code Committee for revisions to certain Code paragraphs to insure that the Code rules are consistent with the reduced design margin of 3.5 and tQiiimplement the necessary changes to the Code rules. This report concludes that a reduction in the present design margin from 4 to 3.5 is justified, based on the improvements in the Code rules and the excellent past experience of vessels built to the Code rules. A. Introduction The ASME Boiler and Pressure Vessel Code was established to provide a uniform set of rules which would result in improved safety of pressure vessels. The first edition was a set of rules published in 1915 Evaluation of Design Margins Section VIII 1 for power boilers which significantly reduced the amount of boiler accidents. 4 The firstASME Code for Pressure vessels was issued as "Rules for the Construction of Unfired Pressure Vessels," Section VIII, 1925 Edition. It had a design margin of 5 on the specified minimum tensile strength. Sections I, III, IV, and VIII now provide rules for construction of power boilers, nuclear power plant components, heating boilers, and unfired pressure vessels. These rules apply to new construction only. The present ASME Code, Section VIII, Division 1 (herein referred to as "the Code") design margins have existed since 1942/1943 when the design margin on the ultimate tensile strength was reduced from 5 to 4 during the World War II, when there was a need to conserve critical materials due to national emergency. Soon after that the hydrostatic test pressure was reduced from 2 to 1.5 times the maximum allowable working pressure. The successful service experience, improved materials, more sophisticated design rules, and improved vessel fabrication practices have now prompted a further review of the ASME Code, Section VIII, Division 1 requirements to see if there is a justification for further reduction in the Code design margins. The purpose of this report is to study the justification for reduction of the design margin on the ultimate tensile strength from the present value of 4 to a proposed value of 3.5. (A change in other design factors is not contemplated at this time). This report will also point out any areas of concern if the design margin on the ultimate tensile strength is reduced. Service experience, test results, and revisions to the Code will be considered in providing recommendations for Code revisions and for follow-up studies. This study is not intended to apply to the parts of Section VIII, Division 1 which cover vessels designed to design margins other than 4 on the specified minimum tensile strength, which excludes vessels fabricated to the following rules: • PartUCI • PartUCD B. Service Experience with Section VIII, Division 1 Vessels The ASME Code, Section VIII, Division 1 includes design rules, fabrication and NDE requirements and specific requirements for pressure relief devices to insure adequate margins of safety. The ASME Code rules apply to new construction only and have no control over the operation of vessel. Although the Code does not address post-construction issues, it does include certain comments and suggestions to avoid service induced problems (Section II, Appendix 6; Section VIII, Appendices NF and HA). These 2 suggestions and recommendations have been added in the later editions of the Code. Although literature contains various articles· on failures of pressure vessels, no comprehensive summary exists on service experience of ASME Code, Section VIII, Division 1 pressure vessels. Also, in some cases information is not available on failures due to ongoing litigation. No attempt has been made here to review all pressure vessel failures. The Bibliography at the end of this report, however, includes several references on failures of pressure vessels. Pressure vessel failures generally fall into the following categories: 1. Failures due to inadequate design or inadequate details (fatigue, etc.) 2. Failures due to poor notch toughness, fabrication defects, or welded repairs. 3. Process or operation related failures (overheating, explosions, etc.) 4. Failures caused by service related degradation or cracking (corrosion, stress corrosion cracking, etc.). 5. Failures from a combination of the causes listed above. The majority of failures seem to be related to poor notch toughness, service degradation, or operation related problems. The incidence of catastrophic failures due to design faults or fabrication faults is low. Very few failures can be attributed. to inadequate design rules in ASME Code. Table 1 lists several failures, available from the published literature, most of which have occurred during hydrostatic testing. Most failures listed in Table 1 have occurred in as-rolled, coarse grain materials with poor notch toughness. Recent revisions to the Code rules recommend hydrostatic testing and require pneumatic testing at temperatures at least 30F above the minimum design metal temperature (UG-99(h) and UG-lOO(c)). This requires a minimum metal temperature at least 30F above the appropriate impact test exemption curve in Fig. UCS-66. Most of the materials listed in Table 1 would be assigned to Curve A. The present Code rules limit the use of most Curve A materials (see UCS-6(b)) and would require the use of better materials or heating the test water to insure an adequate margin of safety during the overload testing. Reference 24 describes a brittle fracture of a 14 ft. diameter, 1.6" thick vessel during stand-by operation below design pressure. This vessel was built of French A-42C-35 material, which is equivalent to SA-201B. This vessel failed after welded repair without PWHT, when the operating pressure was at 65% of design pressure, at about 41°F. Failure initiated at the tip of small surface cracks, from welding residual stresses and localized embrittlement by dynamic strain aging. The failures which have occurred during hydro- WRC Bulletin 435 1954 New 1955 20yrs 1956 75mm(3") A302Gr. B Process vessel 32mm(l1/4'') Equiv. to A 212 Gr. B New 13.6 m (44'--8") dia. x 23.1 m (75'-11") high reactor. (non-PWHT) 32mm(11/4'') A285Gr. C 1956 New 2.13 m (7 ft) dia. CO Converter. (PWHT) 43 mm (1.7") A204Gr. B 1967 New Cylindrical drum (PWHT) 75mm(3") A212 Gr. B 1970 15 yrs 3.657 m (12') l.D. catalytical reformer reactor 51 mm(2") A204Gr. B 1970 New 49.2 m (15') dia. X 42.3 mm (1.625") thk (PWHT) & 8'-10Y2" dia. (non-PWHT) 28 mm(l.l") (as-rolled) A515-1967 Gr. 70 NL Abt. 25 yrs Coke drum NL New 6.1 m (20') dia. x 24.3 m (80') long. (not PWHT) lW' A515-1969 Gr. 70 1972 New 1.524 m (5') 1.D. (PWHT) 178mm (7") A515 Gr. 70 1978 llyrs 305 cm (12') l.D. (PWHT) 7 cm (2.75") A212 Gr. B 1979 23 yrs Heat exchanger NL Al82 Gr. F12 NL New Cylindrical vessel NL A515 Gr. 70 NL NL NL NL NL A204Gr.A A387 Gr. 11Cl.2 1980 17 yrs 2.286 m (7.5') I.D. desulfurizer reactor 85.5 mm (3.25") + 3.5mmclad A204Gr.A 1871 20yrs 4.3 m (14.1') l.D. spherical catalitic reformer 41 mm (1.614") AFNORA!tJ-205 (similar to A201 Gr. B) 1984 16yrs 2.6m (8'-6") I.D. CsfC4 amine absorber vessel 25.4mm(l") SA-516, Gr. 70 Failed at 14°C (57°F) and 16.5 MPa (2390 psig) during hydrostatic test. 30 Failed during hydrostatic test at 16-21°C (61-70°F).so Failed during hydrostatic test at 1.5 times design pressure and 8 to 11°C (46 to 52°F). Failure initiated in the knuckle of the bottom head. 30 Failed during hydrostatic test at 1.5 times design pressure and 7°C (45°F).30 Failed during hydrostatic test at 87% of design pressure and 15°C (60°F).30 Failed during gas test at 90% of design pressure and 11-21°C (52-70°F). 30 Failure occurred during hydrostatic testing at 127% of design pressure and 12°C (54°F). Failure occurred from a preexisting defect at a fillet weld attaching manway reinforcing plate to shell. 22 Failure occurred during hydrostatic testing at 315 psig (abt 94% of max test pressure) and 4 °C (40°F) after welded repairs and PWHT. 29 Failure occurred during hydrostatic test at near max. test pressure of260 psig. Test water temp. not known. Failure initiated at toe of fillet weld between manhole reinforcing plate and vessel shell. Failure occurred during hydrostatic test at 131 % of design pressure at 18°C (64°F).30 Failure occurred during hydrostatic testing, after repairs, at 880 psig internal pressure and 25°C (77°F). (Design pressure was 650 psig). Failure initiated from a partial penetration weld between two halves of steam dome reinforcing pad. 21 Vessel failed during hydrostatic test at 154% of design pressure. 30 Vessel failed during hydrostatic test at 13.8 Mpa (2000 psig) and l6°C (61°F).23 Vessel failed during hydrostatic test at 10°C (50°F).30 Vessel failed during inert gas test at 85% design pressure and ambient temperature.so Vessel failed during standby service at about 65% of design pressure at 5°C (41°F).24 Vessel failed during operation (at about 200 psi and 100°F) due hydrogen induced cracks inhardHAZ =not listed. Evaluation of Design Margins Section VIII 3 static testing are by no means limited to coarse grain steels. Several liquid spheres failed in a brittle manner during hydrostatic testing in Europe during 1960-1970's. 25 At least four of these are known to be of fine grain, fully killed, normalized steels. The failures of these spheres are attributed to poor details and improper fabrication practices for the materials used. These spheres were designed to stresses permitted in European codes, but no detailed reports are available to the authors on these failures. C. Failure Modes Considered in the Code Rules In 1987-1989 the ASME Subgroup on Design Analysis (SGDA) performed a study to determine the failure modes considered in the sections of the Code and the intended safety margins for a given failure mode in each section of the Code. Eight failure modes were identified and discussed in the February 1989 report "Summary of the Design Analysis Factors Inherent in the Established Failure Modes of the ASME boiler and Pressure Vessel code," prepared by Subgroup Design Analysis. The following is a brief description of each failure mode considered in Section VIII, Division 1. A description of design margins explicitly introduced into the Code design rules is included in Appendix 1, Section II, Part D. 1. Ductile Rupture The following general design margins have been established to provide an adequate margin of safety against ductile rupture below the creep range: • SulS;:::: 4 • SylS ;:::: 1.11 (for some non-ferrous alloys and austenitic materials) • Sy!S;:::: 1.5 (for all other materials) S = Allowable design stress Sy = Specified minimum yield strength Su = Specified minimum tensile strength The following references in the Code give the allowable stress values for Section VIII, Division 1 and the basis for the allowable stress values. • Section VIII, Division 1, Appendix P, Basis for Establishing Allowable Stress Values. • Section II, Part D: Tables lA and lB. • Section II, Part D, Appendix 1, Nonmandatory Basis for Establishing Stress Values in Tables lAand lB. The maximum allowable stress is the lowest value obtained from the section VIII, Division 1 criteria at the specified design temperature. For structural materials, when used for pressure retention applications, and for bars and shapes used as shell stiffeners or stay bolts, the basis for allowable stresses is subject to a quality factor of 0.92 (see Paragraph 1-lOO(c) of Appendix 1 of Section II, Part D). This quality factor is based on judgment and 4 experience for present design margins. With a lowering of the design margin, it is recommended that the Code Committee review this factor. The basis for establishing allowables for Parts ULT, UCI, and UCD of the Code are spelled out in Appendix P. The design margins for ULT-23 materials are generally the same as for wrought or cast materials. It is assumed that any reduction in design margins agreed upon by the Code Committee will be applicable to these materials. However, the design margins for UCI-23 and UCD-23 materials are different from that of wrought or cast materials, and it is assumed that these factors will remain unchanged. The allowables for welded pipe or tube are generally 85% of the value obtained by using the basis for wrought or cast materials (see Table 1-100 of Appendix 1 of Section II, Part D). It is assumed that this factor will be retained and will be applicable to any revised design margins. It is assumed that the design margin for bolting materials will also be revised to remain consistent with other materials. The present Code uses generally the same basis for establishing bolting allowables, i.e. 114 of ultimate or % of yield, except for materials whose strength has been enhanced by heat treatment or by strain hardening (see 2-120 of Appendix 2 of Section II, Part D). Section E of this report comments on the burst tests performed on SA240 Type 304, SA516 Gr. 70 and SA517 Gr. F steel test vessels at University of Kansas. 12 Based on these tests, it is unlikely that the true margin of safety against ductile rupture would be affected by lowering the design margin from 4 to 3.5 on ultimate tensile strength. 2. Brittle Fracture This failure mode is addressed in the Code by rules for design, the notch toughness requirements in UG-84, UCS-66, UCS-67 and UCS-68, fabrication requirements, and hydrostatic/pneumatic testing requirements. The following are examples of Code rules which guard against brittle fracture: UG-16 through UG-35 (design), UG-84 (Charpy impact tests), UG-85 (Heat treatment), UG-95 thru UG-97 (examinations and inspections during fabrication), UG-99 thru UG-100 (Hydrostatic and pneumatic tests), UG-125 thru UG-136 (Pressure relief devices). Part UW (Requirements for pressure vessels fabricated by welding). UCS-56 (Postweld heat treatment), UCS-65 thru UCS-68 (Low temperature operation), UCS-75 thru UCS-85 (Fabrication). UHA-32 (Postweld heat treatment), UHA-50 thru UHA-52 (Inspection and tests), Appendix HA (Suggestions on selection and heat treatment of high alloy steels). UHT-5 & UHT-6 (Materials), UHT-17 (Weldedjoints), UHT-18 (Nozzles), UHT-56 (Postweld heat treat- WRC Bulletin 435 ment), UHT-57 (Examination), UHT-75 thru UHT-86 (Fabrication). Appendices 4, 6, 7, 8, 10, 12 (requirements for non-destructive examination and quality control). Appendix R (Preheating). The Section VIII, Division 1 toughness rules are based on fracture mechanics considerations and extensive good experience with ASME Code vessels. It is not possible, however, to establish a single margin of safety against brittle fracture in the Code rules because brittle fracture depends on three factors, the applied stress, the flaw dimensions, and the available toughness of the material. In fracture mechanics calculations, the value of the stress intensity, Kr, is based on the applied stress and dimensions of the flaw. The calculated Kr must be less than the critical fracture toughness parameter Kic (or Krd, depending on the loading rate) to avoid brittle fracture. Also, consideration must be given to both primary and secondary stresses to evaluate maximum stress intensity, Kr, in the structure. Work by Dr. Sumio Yukawa 16 attempts to evaluate the Code margins for resistance to brittle fracture in carbon and low alloy steels based on: 1. Critical vs. permissible flaw length for a given stress, and 2. Material fracture toughness vs. applied stress intensity factor, Kr, for a given flaw size. The report by Dr. Yukawa gives the following results: • Calculated critical flaw length vs. maximum Code permissible flaw size of%" for RT examination, for various flaw depth/thickness (alt) ratios and applied stress/yield strength ratios (SR). The size margin for SR = 1.0 and alt = 0.5 is about 2 Yz in 1-3 in. thick material with 38 ksi minimum yield strength (and higher for lower alt ratios and lower stress ratios). i • Available fracture toughness (based on mihimum required toughness in the Code) vs. the applied stress intensity, K, for a flaw with a length of%" (maximum permitted in UW-51 for RT examination). This evaluation indicates a minimum margin of about 1.5 for 1 and 3" thicknesses (and higher values for lower stress ratios) for a material with a minimum yield strength of38 ksi and stress ratio (SR) of 1.0. • This study also concludes that the relief from notch toughness testing as permitted in Figure UCS-66.1 does not compromise these margins. It should be noted that any indication characterized as a crack or zone of incomplete fusion or penetration in welded joints is unacceptable by the Code rules CUW-51(b)(l) and UW-52(c)(l)). A further review of the Code margins for resistance to brittle fracture is being planned for the Phase 2 study. This will include the effect of higher design stresses (i.e. lower margins against ductile rupture) on margins against brittle fracture for Part UCS and Part UHT materials. It is, however, the opinion of the authors that the safety of Section VIII, Division 1 vessels would not be jeopardized by lowering the design margin from 4 to 3.5, because of the following considerations: 1. The Section VIII, Division 1 toughness rules (introduced in 1987). 2. Enhanced notch toughness resulting from improved steel-making practices (see Sect. D.l). 3. Improved welding practices. 4. The successful experience with Section VIII, Division 2 vessels designed to higher stresses but similar toughness requirementp as the present Division 1 vessels. 3. Fatigue Section VIII, Division 1 presently has no rules for fatigue analysis. The cyclic use for Division 1 vessels is assumed to be low. Paragraph UG-22 requires consideration of cyclic loads, but the Code does not indicate what is a significant number of cycles for which fatigue analysis should be performed, therefore, the designer is referred to U-2(g). The exemption from fatigue rules of Section VIII, Div. 2 provide a quantitative estimate of the number of cycles of pressure and temperature loading that a vessel can safely tolerate, when designed to a certain stress limit. In Phase 2, calculations will be included to quantify the effect oflower design margins. 4. Incremental Collapse Incremental collapse is generally caused by ratchetting which is inelastic strain accumulation due to cyclic loading. The Code does not provide explicit rules for this mode of failure. Design and fabrication rules of the Code are intended to minimize the causes of ratchetting. If a vessel is in a cyclic service and a fatigue analysis is performed, the required limit on the range of primary plus secondary stress intensity (required to be limited to two times yield strength by Section VIII, Div. 2 and Section III) will guard against ratchetting and incremental collapse, therefore, the safety of Section VIII, Division 1 vessels would not be adversely affected by lowering the design margin to 3.5. 5. Elastic Instability The C?~ rules on elastic instability are based on geometric parameters and are independent of the design stress limits. Any change in the design margin on ultimate strength will not affect the buckling rules. (A PVRC study has proposed alternative rules to the present Code rules on buckling. The report has been published as a WRC Bulletin 406). For Code basis in developing external pressure curves see Appendix 3 of Section II, Part D. 6. Plastic Instability As the R/T ratio of shells get smaller, the failure mode changes from elastic instability to plastic Evaluation of Design Margins Section VIII 5 instability. The present Code rules for design of formed heads are generally based on plastic collapse. The plastic collapse is a strain induced failure, even though the Code rules are based on stress allowables. A reduction in design margins on ultimate strength will reduce the margins for these components. But these rules are generally regarded as very conservative and, therefore, it is unlikely that decreasing the design margin from 4 to 3.5 on UTS would affect the safety of the vessel. The Code external pressure curves, at the upper end, are limited to % of yield strength and these curves will not be affected by a change in design margins. The calculated allowable compressive stresses are not allowed to exceed the basic Code stress allowables. 7. Creep Rupture The Code allowable stresses at temperatures in the creep range provide a reasonable margin of safety against creep rupture failure. The allowable stresses at temperatures in the creep range are based on the lower of the following: • 100% of average stress to produce a creep rate of 0.01 %/1000 hrs. • 67% of the average stress to cause rupture at the end of 100,000 hrs. • 80% of the minimum stress to cause rupture at the end of 100,000 hrs. These limits will not be affected by any reduction in design margin on ultimate tensile strength, however, the transition from time independent properties to time dependent properties will occur at a lower temperature. 8. Creep Buckling The temperature limits on the Code buckling curves are intended to be limited to the temperature at which creep becomes a consideration for the material. As such, the present Code does not intend to address creep buckling. However, some of the temperatures on Code external pressure curves do extend into the creep range and the Code Committee is looking into this (PVRC has been working on recommendations for creep buckling rules to be recommended to the Code Committee). Non-Mandatory Appendices. As the needs developed over the years, many new materials and Parts were added to Section VIII. The Requirements for Pressure Vessels Constructed of Carbon and Low Alloy Steels are now included in Part UCS. 1. Materials Table 2 lists the carbon and low alloy steel plate specifications which were included in the 1950 edition of Section VIII, Table UG-23. Table 3 lists the carbon steel and low alloy steel plate specifications included in Table lAofthe 1992 edition of Section II, Part D, for comparison. It is obvious that there have significant changes to the type of materials permitted for construction of Section VIII vessels. The following addresses some of the changes to the materials since the 1950 edition of Section VIII: SA-30. SA-30 was as-rolled low strength carbon steel with specified carbon and manganese contents, but had no requirements for deoxidation. This specification has been withdrawn. SA-129. This was a low strength carbon steel. This specification has been withdrawn. SA-201. SA-201, Grades A and B were low strength, fully killed carbon steels with 0.80% maximum manganese content. SA-201 had no grain size requirements, therefore both grades could be produced to fine grain or coarse grain practice. Plates under 2" thick were typically produced as-rolled. Plates over 2" thick were required to be normalized. This was a commonly used steel for pressure vessel construction. This specification has now been voided. SA-201B has been incorporated in SA-515, as grade 60. SA-212. This steel was available in grades A (65 ksi min. tensile strength) and Grade B (70 ksi min. tensile strength). These were fully killed steels with a maximum manganese content of 0.90%. Grade A had a maximum carbon content of 0.31 % in thicknesses up to and including l" and 0.33% for Table 2-Carbon and Low Alloy Steel Plate Specifications Included in th e1950 Edition of ASME Section VIII 6 Min. Tensile SA-283 SA-36 SA-285 SA-299 SA-455 SA-5I5 SA-5I6 SA-662 SA-537 SA-737 SA-738 SA-84I (CC2130-I) SA-6I2 SA-225 SA-204 SA-302 SA-30 SA-I29 SA-20I SA-2I2 SA-285 SA-299 SA-30I SAI-302 A-283 SA-202 SA-203 SA-204 SA-225 WRC Bulletin 435 Flg,AFbx,B Fbx A,B A,B A,B,C A,B A,B,C,D A,B A,B,C A,B,C A,B Carbon 55, 42, 48, resp. Carbon Carbon C-Si Carbon C-Mn-Si I\i,iCr-YzMo Mn-YzMo Carbon 44 55, 60, resp. 65, 70, resp. 45, 50, 55, resp. 75 70 75, 80, resp. 45, 50, 55, 60, resp. 75, 85, resp. 65, 70, 75 resp. 65, 70, 75, resp. 70, 75, resp. A,B,C,D A B,C Carbon C-Mn-Si Carbon C-Mn-Si C-Mn C-Si C-Si C-Mn-Si C-Mn-Si C-Mn-Si C-Mn-Si C-Mn-Si C-Mn-Si C-Mn-Si-Cb C-Mn-Si-V C-Mn-Si C-Mn-Si c C-Mn-Si Mn-YzNi-V A,B,C 60, 65 70 55, 60, 65 70 A,B c I 2, 3 B c C-YzMo C-YzMo A B,C A B c SA-533 SA-387 SA-202 SA-542 (CCI961-I) SA-542CCCI960-4) SA-542(CC2098-2) SA-832 SA-517 SA-203 SA-543 SA-353 SA-553 SA-645 D. Improvements to the Code Since 1940's There have been numerous changes and improvements in the Code rules since the present design margins on ductile rupture were established in 1942/3. Certain current Code requirements are compared to those in the 1950 edition of Section VIII in this report for comparison. The 1950 edition of Section VIII consisted of about 280 pages (5114 X 7%") and included Parts UG (General Requirements for Vessels of Carbon and Low Alloy Steels), UW (Requirements for Welded Vessels of Carbon and Low Alloy Steel), UR (Requirements for Riveted Vessels of Carbon and Low Alloy Steel) and Mandatory and Table 3-Carbon and Low Alloy Steel Plate Specifications Included in 1992 Edition of ASME Code Section II, Part D and Section VIII Code Cases D A, Cl.I & Cl.2 A,Cl.3 B, Cl.I B,Cl.2 B,Cl.3 C, Cl.I & Cl.2 C,Cl.3 2, Cl.I 2,Cl.2 I2, Cl.I & Cl.2 11, Cl.I & Cl.2 22, Cl.I & Cl.2 21, Cl.I & Cl.2 5, Cl.I &Cl.2 9I A,B C,Cl.4 B, Cl.4 D,Cl.4 A B E J p A,B D,E,F C, Cl.I, 2 & 3 B, Cl.I, 2 & 3 I, II Mn-YzMo Mn-YzMo Mn-Y2Mo-%Ni Mn-%Mo-%Ni Mn-Y2Mo Mn-YzMo Mn-YzMo-YzNi Mn-YzMo-YzNi Mn-YzMo-YzNi Mn-YzMo-%Ni Mn-YzMo-%Ni YzCr-YzMo YzCr-YzMo ICr-YzMo I;i.iCr-YzMo 2;/.iCr-IMo 3Cr-1Mo 5Cr-YzMo 9Cr-1Mo-V YzCr-IY-iMn-Si 3Cr-1Mo-;i.iV-Ti-B 2;1.iCr-lMo 2;1.iCr-IMo-;i.iV 3Cr-IMo-\4V-Ti-B YzCr-;/.i~o-Si YzCr-YsMo-Si Io/.Cr-YzMo-Ti C-YzMo l;i.iNi-lCr-YzMo 2YzNi 3YzNi 2%Ni-l VzCr-YzMo 3Cr-lo/4Cr-%Mo 9Ni 9Ni,8Ni 5Ni-;i.iMo I Gr. I I Gr. I I Gr. I I Gr. 2 I Gr. 2 I Gr. I I Gr. 2 I Gr. I 1 Gr.2 1 Gr. I lGr. 2 1Gr.2 1Gr.3 I Gr. 2 I Gr. 3 I Gr. 2 I Gr. 2 I Gr. 2 IOC lOAGr. I 3Gr. I 3Gr.2 3Gr. 2 3Gr. 3 3Gr. 3 3 Gr. 3 3Gr. 3 llAGr. 3 3 Gr. 3 3 Gr. 3 llAGr. 4 3 Gr. 3 11Gr.4 3Gr. I 3 Gr. 2 4Gr. I 4Gr. l 5AGr. I 5AGr. l 5B Gr. I 5B Gr. 2 4Gr.1 5C Gr. I 5C Gr. I llB Gr. 1 llB Gr. 4 llB Gr. 2 llB Gr. 6 11B Gr. 8 9AGr. l 9B Gr. I llAGr. 5 llAGr. 5 llAGr. I llAGr. I llAGr. 2 45, 50, 55, 60, resp. 58 45, 50, 55, resp. 75 70 60, 65, resp. 70 55, 60, 65, resp. 70 58, 65, resp. 70 70(65)* 80(75,70)* 70 80 70 85, 80(75,70)* 70 83(8I)* 105 65 70, 75, resp. 75 80 80 80 80, 90, resp. 100 80 90 100 80, 90, resp. 100 55 70 55, 65, resp. 60, 75, resp. 60, 75, resp. 60, 70, resp. 60, 75, resp. 85 75, 85, resp. 85 85 85 85 115 115 115 115 115 65, 70, resp. 65, 70, 80, resp. 105, 115, 90, resp. 105, 115, 90, resp. 100 IOO 95 *Varies with thickness %Cr-I\4Mn-Si 2%Ni C-YzMo Mn-YzNi-V plates over l" thick. These grades were normally produced to coarse grain practice. Plates up to and including 2" thick were typically produced asrolled; thicker plates were required to be normalized. These were commonly used grades for pressure vessel construction. This specification has now been voided and the steels included in SA-515 as grades 65 and 70. SA-283. These are low strength structural steels without any specific deoxidation requirements. Grades A (45 ksi min. UTS), B (50 ksi min. UTS), C (55 ksi min. UTS) and D (60 ksi min. UTS) are still permitted for pressure vessel construction, but their use is limited to%" maximum thickness for pressure parts and subject to the toughness requirements in UCS-66. Evaluation of Design Margins Section VIII 7 SA-285. Grades A (45 ksi minimum UTS), B (50 ksi min. UTS) and C (55 ksi min. UTS) are low strength carbon steels with 0.90 max. manganese content and no requirements for deoxidation. These grades are still permitted for pressure vessel construction, subject to the notch toughness requirements in UCS-66. SA-299. This is a fully killed C-Mn steel with 75 ksi minimum UTS and can be produced to fine grain or coarse grain practice. Plates 2" thick and thinner can be produced as-rolled. Plates over 2" thick must be normalized. This steel is still used for Code pressure vessels, but it's use now is restricted to Curve A steels in Fig. UCS-66 for notch toughness consideration (unless it is produced to fine grain practice, in which case it is considered Curve B steel). SA-300. This specification included toughness requirements for normalized steel plates for pressure vessels. The impact tests were to meet 15 ft-lbs minimum average, using a Charpy type, keyhole (or milled "U") notch specimen. This specification has now been voided, as the impact test requirements are now included in the individual materials specification and in the ASME code. All impact tests now must be performed using V-notch type Charpy specimen conforming to Fig. 11 in ASTM A370 (which is a more severe test than a keyhole type specimen). SA-202. Gr. A (75 ksi min. UTS) and Gr. B (85 ksi min. UTS) are Yz Cr-1114 Mn-Si fully killed low alloy steels which can be supplied in as-rolled condition or normalized. These steels are still permitted for Code vessels, but their use is restricted Curve A or Curve B steels (as applicable) for notch toughness considerations. SA-203. Grades A and B are fully killed, normalized, 2%% Ni steels with good notch toughness and are still included in the present edition of Section II, Part D for Code vessels. Grade Chas been eliminated, but several new 3%% Ni grades have been added. SA-204. These are fully killed C-Yz Mo steels and are available in Grades A, B and C, depending on strength. These steels can be produced as coarse grain or fine grain steels and do not need to be heat treated in thiclmesses up to and including 11/z". These steels are still permitted for Code vessels, but have poor notch toughness in the as-rolled condition, and their use is now restricted to Curve A or Curve B (as applicable) for notch toughness considerations. SA-301. These 1114 Cr-Yz Mo and 1 Cr-Yz Mo low alloy steels have now been replaced with SA-387 Grades 11and12. SA-302. Grades A and B are C-Yz Mo steels, similar to SA-204, but have higher manganese contents and higher strength. SA-302 Grades A and B are now subject to the same notch toughness considerations in UCS-66 as the SA-204 steels. 8 Many more steels have now been approved for Section VIII Division 1 vessels than those listed in the 1950 edition oftheASME Code. Almost all of the newer steels are fully killed, and many are made to fine grain practice and heat treated. This results in better notch toughness and reduces the risk of brittle fracture and improves safety. An exception is SA-515 which is produced to coarse grain practice. The SA-515 grades are permitted for use at design metal temperatures up to and including 1000°F and have the same allowable stresses as the comparable SA516 grades (made to fine grain practice). The SA-515 grades, however, have poorer notch toughness because of the coarse grain practice. Grades 65 and 70 are assigned to impact test exemption Curve A and grade 60 to Curve Bin Figure UCS-66. All SA-516 grades are fully killed, made to fine grain practice and are normalized when the thickness exceeds 11/z". As-rolled grades 65 and 70 are assigned to impact test exemption Curve B and as-rolled grades 55 and 60 to Curve C. When normalized, all SA-516 grades are assigned to Curve D in Fig. UCS-66 and can be used at lower design metal temperatures without impact testing. Another example is SA-537, Classes 1, 2 and 3. These are fine grain, fully killed steels which are heat treated in all thicknesses, and therefore suitable for lower metal design temperatures. Part UHT includes requirements for pressure vessels constructed of ferritic steels with tensile properties enhanced by heat treatment. Most of the steels included in Part UHT were developed since the 1950 edition of the Code. Some of the steels in Part UHT are unique in that they were subjected to large scale low temperature vessel tests to demonstrate their suitability for Code construction of pressure vessels. Four test vessels of SA-517 Gr. F (USS Tl steel) were burst tested at about -45°F, and additional four vessels were pressurized and subjected to falling weight drop-impact tests at temperatures from -33°F to -41°F. Similar tests were performed on SA-553/SA-353 (9% Ni) steel test vessels at - 320°F to justify the use of these steels at cryogenic temperatures without post-weld heat treatment. There have been a number of improvements in steel making practices since the 1950's. For example, most steels are now produced by the continuous casting process. This requires the use of killed steels with aluminum or aluminum plus silicon additions to achieve full deoxidation of the steel and insure good castability. Continuous casting also helps to achieve more uniform properties in the plates. Improvements have also been made to slag processing, refractories, secondary refining, temperature control, rolling mills, heat treatment, quality control, etc., which has resulted in cleaner steels with lower carbon, sulfur and phosphorus contents and better quality materials. For instance, in recent years ASTM and ASME specifications have reduced the WRC Bulletin 435 maximum phosphorus and sulfur contents by 0.01 % in several pressure vessel steels. It is well recognized that reducing carbon and sulfur contents improves notch toughness (See Figures 1-3).All these improvements have a beneficial effect on notch toughness, however, none of these improvements have been reflected so far in the Code design margins. 200 125 150 .0 IOOf >-" 75 IE >-" (,!) 100 a: w w 50 ~ z w 50 2. Design The design rules of the Code have been extensively expanded and improved since the 1940's. The additions and revisions to these rules are too numerous to list here. The following is a brief description of some of the more significant improvements to the rules which are judged to add to the safety of vessels designed and constructed to these rules. a. Design of cylinders for internal pressure: A criteria for longitudinal stress has been added [UG-27(c)(2)] and formulas for thick shells have been provided [Appendix 1]. b. Design ofspherical shells for internal pressure: The value of "0.2P" has been introduced for calculating design thickness [UG-27(d)]. For thicker shells, this will result in a measurable increase in design thickness. Formulas for thick shells have been provided [Appendix l]. c. Joint efficiencies: The 1949 Edition of the Code allowed joint efficiencies of 90% and 80% for longitudinal joints with full RT or no RT. However, the use of these values were contingent upon many other welding and testing requirements. Table 4 lists the maximum joint efficiencies in the 1950 ASME Code, Section VIII and in the 1992 edition of Section VIII, Division 1. The present rules are so extensively different that it is difficult to make a direct comparison. However, it is obvious that today's rules are much more detailed and sophisticated. d. Design of formed heads: The formula for design ofhemi-spherical heads was almost the same (it did not include the 0.2P value). The only desiifn formula for "dished heads" was that for spherically dished covers of present Code. 1- 6 Extensive formulas and rules for design of other types of formed heads 25 0 0 .018 .024 SULFUR,% Fig. 2-Effect of sulfur content on the shelf energy of straightaway rolled carbon steel plate33 have been added. Limitations on geometric parameters have been added. e. Design of flat heads: The basic design formula was the same as that of present Code, but did not have the joint efficiency included. Also the values of "C" in present Code cover more cases and are generally more conservative. f. Design of bolted flanged connections: The rules were basically the same as those in the present Code, except a great deal of details have been added over the years. g. Design ofopenings: The rules allowed considerably larger openings than those of today's Code [UG-36(c) 3 ] to be exempt from reinforcement requirements. There were no spacing limitations for multiple unreinforced openings (For calculation of ligament efficiencies, reference was made to Section I). For those openings not exempt, the reinforcement requirements were similar to today's rules. Full pressure area had to be replaced, but the limits of reinforcement were somewhat different. A great deal of detail and refinements have been added over the 60 200 - - - - - - - - 0.11%Carbon ..., >-' (!) a: w z w 1-- 150 0.20% Carbon 100 0.31% Carbon (..) <l: a. ~ 20 50 -100 TEST TEMPERATURE, °C. Fig. 1-Effect of carbon content on transition curves of ferriteprearlite steels32 -50 0 +50 TEMPERATURE, "F +100 Fig. 3-Effect of sulfur content 0n transverse impact properties (Bethlehem Steel research data on HSLA steels)31 Evaluation of Design Margins Section VIII 9 Table 4-Maximum Allowable Joint Efficiencies for Butt Welded Joints 1. From 1950 ASME Code, Section VIII Table UW-12: Spot RT Limitations Type of Joint Thermally PWHT'd None Single welded butt joint, with backing strip. Longitudinal joints not over lYi'' thick. (No thickness limitation on circumferential joints). Circumferential joints only, not over%" (0.80) 2. From 1992ASME Code, Section VIII, Division 1, Table UW-12: x Max. Joint Efficiency 0.80 x Double welded butt joint. without Full RT x x x x 0.85 0.90 0.95 0.70 Joint Description Limitations Joint Category Full RT Spot RT None Butt joints as attained by doublewelding or by other means which will obtain the same quality of deposited weld metal on inside and outside weld surfaces to agree with the requirements ofUW-35. Welds using metal backing strips which remain in place are excluded. Single-welded butt joint with backing strip other than those included above. None A,B,C,D 1.0 0.85 0.70 Single-welded butt joint without use of backing strip (a) None except as in (b) below. (b) Circumferential butt joints with one plate offset; see UW-13(b)(4) and Fig. UW-13.1 and sketch (k). Circumferential butt joints only, not over%" thick and not over 24" outside dia. years, which will result in a better and more conservative design. h. New rules: In addition to improvements to the old rules, a great many new design rules have been added to improve the safety and quality of vessels. A lot of the recent design rules, although put into design by formula approach for easy application, are based on modern analytical methods. If a certain geometry or loading is not covered by Code rules, paragraph U-2(g) may be used to analyze. For this purpose many advanced computational techniques are available, which were not available when the present design factors were incorporated. Some of the design rules added to the Code since the 1940's are cone-cylinder junction design rules, ligament design rules, rules for multiple openings, requirements for minimum nozzle neck thickness, design rules for braced and stayed surfaces, and rules for large openings in cylindrical shells. A number of appendices have been added to provide design rules for specific components such as jacketed vessels, vessels of non-circular cross section, integral fiat heads with a large opening, dimpled or embossed assemblies, integrally forged vessels, clamp connections, and expansion joints. These specific and detailed rules can only contribute to quality and uniformity of vessels built to the Code. The result of these improvements is that better and more refined design rules have been established, which has improved the safety of Division 1 vessels. 3. Fabrication There have been several improvements in the fabrication requirements of Section VIII pressure 10 A,B,C,D 0.90 0.80 0.65 A,B,C 0.90 0.80 0.65 A,B,C NA NA 0.60 vessels since the 1940's. These involve mainly the requirements for welding, inspection, postweld heat treatment and quality control. The following changes since the 1950 edition oftheASME Code are believed to have improved the quality of Section VIII pressure vessels and margins of safety: Forming: No specific limitations were included in the early editions of Section VIII (e.g. 1950). UCS-79(d) now limits the maximum forming strains to 5% without heat treatment if the vessel will contain lethal substances or the material is impact tested. UHT-79(a) 1 limits the maximum forming strains to 5% for high strength materials without heat treatment. Joint Efficiencies for Welded Joints: The joint efficiencies in Table UW-12 have been further refined. The maximum joint efficiencies are now based on the type of joint and the degree of radiographic examination (and not on PWHT). Table 4 provides a brief overview of the maximum joint efficiencies in Tables UW-12 of the 1950 Code, Section VIII and the 1992 edition of Section VIII, Division 1 for double butt welded joints and single butt welded joints with backing strips. For Type 1 joints with 100% RT the maximum joint efficiency has been increased to 1.0 (presumably because of improved NDE techniques). For all types ofjoints without any RT examination, the joint efficiencies in the 1950 Edition have been decreased to lower values. The present UW-35 includes more detailed requirements for the finished longitudinal and circumferential joints than the earlier editions of the Code. WRC Bulletin 435 UW-35(a) now specifically states that "Butt welded joints shall have complete penetration and fusion," which insures better quality welds and, therefore, reliability of vessels. Attachment Welds and Welded Connections: UW13 has been expanded significantly from the early editions of the Code. New and more detailed requirements have been added in several areas, e.g.: • Limitations and specific details for head-tocylindrical shell transitions, as shown in Fig. UW-13.1 sketches (1), (m), (n), and (o). • Specific requirements for welded unstayed fiat heads to shells in UW-13 (e) and (f), and as shown in Fig. UW-13.2 and Fig. UW-13.3. The requirements for attachment welds at openings have been significantly expanded since the early editions of the Code. Specific requirements are now included in UW-15(c) for groove welds and fillet welds. Entire UW-16 and Fig. UW-16.1 have been added since the 1950 edition of Section VIII and provide very specific requirements for attachment welds at openings. (Only some of the acceptable types of welded nozzles in Fig. UW-16.1 were included in Fig. UW-15 of the 1950 edition of the Code). A number of fabrication tolerances have been added over the years which improve the quality of the vessels. Up to 1949 edition of the Code, the out-of-roundness was limited to a value similar to present Fig. UG-80.1. In the 1950 Code, the out-ofroundness tolerances and external pressure tolerance were correctly differentiated and the limit of 1% on out-of-roundness was introduced. It was also in this Edition that tolerances for formed heads were introduced. 4. Welding Significant advances have been made over the last 40-50 years in improving the welding processes and procedures. All materials are now assigned Pnumbers in Section IX of the ASME Code. The welding procedure qualifications requirements in Section IX and several other requirements in the other book sections for fabrication are related to P-numbers. Recommended good practice for preheating various steels is now included in Appendix R of Section VIII, Division 1. Few welding electrode specifications existed at the time of the early editions of the Code. New electrodes and welding processes have been introduced since that time and are now covered by AWS specifications. Standardized specifications now reduce the factors of uncertainty and improve reliability of vessels. One of the more significant improvements has been the use of low hydrogen electrodes which started in 1950's. This reduced the risk of cracking and improved notch toughness of welds. Another recent improvement is moisture resistant electrodes. Much more knowledge exists today about welding, weldability and preheat requirements for various steels than at the time of the early editions of the Code. 5. Postweld Heat Treatment (PWHT) The early editions of the Code (e.g. 1950 edition of Section VIII) included all PWHT requirements in UW-40. The minimum PWHT temperature for all materials was llOOF (with longer times at lower temperatures). The minimum holding time was 1 hr/in., regardless of thickness. Later editions of Section VIII have expanded the PWHT requirements. UW-40 (Procedures for Postweld Heat Treatment) includes much of the same requirements as the early editions of the Code, but provides a better definition of the nominal thicknesses of welded joints and the depth of repairs. The minimum PWHT temperatures, holding times, and exemptions for various classes of materials are included in UCS-56, UHT-56, UHA-32, and UNF-56, depending on the P-number of that material. This is an improvement over the early editions of the Code as the present requirements consider the chemical composition of the various materials. The current Code rules in UCS-56(e) and (f) also include specific requirements for welded repairs after PWHT. Other improvements are: • Heat treatment verification test requirements in UCS-85 and UHT-81 on test specimens which are subjected to the same heat treatments as the vessel material during the fabrication (except where exempted by the Code rules). (This was not required in the early editions of the Code). • Ongoing activities in the Code committees to develop better rules for local PWHT and allowable temperature gradients. • Ongoing activities in the Code committees to reexamine minimum PWHT temperatures and holding times for certain types of materials. The present requirements insure that no harmful distortions or excessive stresses occur in the heat treated vessel and that the required properties are met in the vessel materials after all heat treatment needed for fabrication. 6. NDE Several important changes to the Code have increased the NDE requirements since the early editions ofth~ode: UW-50 now requires the examination of the full length of all welds around openings and all attachment welds having a throat thickness greater than 14" for all vessels to be pneumatically tested. UW-51(a) and UW-51(b) require radiographic procedures to be in accordance with Article 2 of Section V of the Code. The summer 1973 Adenda incorporated several important changes to the RT requirements: • Fine grain film must be used. • No :fluorescent screens are allowed. Evaluation of Design Margins Section VIII 11 • The density limits are more restrictive. • The penetrameters changed in thickness and in hole size. • A detailed written procedure must be qualified in accordance with SNT-TC-lA. The Winter 1975 Adenda included some changes in Article 2 of Section V, which made the radiographic requirements somewhat less restrictive than the summer 1973 Adenda. UW-51(a)2 requires personnel performing and evaluating radiographic examinations to be qualified and certified in accordance with a written practice, in accordance with SNT-TC-lA. UW-51(c)2 adds a requirement that provisions for training, experience, qualification and certification of personnel shall be described in the Manufacturer's Quality Control System U-2(h)). UW-51(b) 1 and UW-52(c) 1 specify that welds in which indications are characterized as cracks or zones of incomplete fusion or penetration shall be unacceptable (regardless of length). The more recent editions of Section VIII, Division 1 now include mandatory appendices which include specific requirements for certification ofNDE personnel, evaluation of indications and acceptance standards. These are: • Appendix 6, Methods for Magnetic Particle Examination. • Appendix 8, Methods for Liquid Penetrant Examination. • Appendix 12, Ultrasonic Examination of Welds. • Appendix 4, Rounded Indication Charts Acceptance. Standard for Radiographically Determined Rounded Indications in Welds. All these changes contribute to improved quality of welds and improved safety of vessels. 7. Brittle Fracture Notch toughness requirements were included in the Code some time in 1940's. The 1950 Code included the following requirements for carbon and low alloy steels: • No impact testing required on any material for use at temperatures of-20°F and above. • No impact testing required when the operating cycle is such that the pressure at -20°F is less than one fifth of the maximum allowable working pressure at 100°F. • No impact testing is required on rivets, nuts, or material less than 0.098 in. thickness. • When impact testing is required, the impact values shall meet 15 ft-lbs, minimum average on Charpy type, keyhole or U-notch specimens. Keyhole and U-notch impact test specimens have a larger notch root radius than the V-notch test specimens, therefore, these are less severe impact tests than the Charpy V-notch test in the current edition of the Code. 12 One of the most significant revisions to Section VIII, Division 1 were in 1987, when the Code published the new toughness rules for carbon and low alloy steels. These rules are based on linear elastic fracture mechanics and the extensive good experience with the existing pressure vessels. The main features in the new toughness rules in Part UCS are: • Fig. UCS-66 impact test exemption curves. These curves take into account the material specification (deoxidation, chemical composition, heat treatment, etc.), the material thickness and the minimum design metal temperature. The less notch tough materials are included in Curves A and B and the more notch tough materials in curves C and D. Unless otherwise exempted by the Code rules, impact testing is required for a combination of minimum design metal temperature and thickness which is below the curve assigned to the material in question (except where specifically exempted by the Code rules). If the minimum design metal temperature and thickness combination is above the applicable curve, impact testing is not required, except as required by UCS-67(a) 2 for weld metal. Impact testing is required for all welded thicknesses over 4" and all unwelded thicknesses over 611 • • Impact testing requirements in UG-84. Fig. UG-84.1 requires increased Charpy V-notch energy values based on thickness and yield strength. These Charpy V-notch values are based on fracture mechanics consideration and experience. • Impact testing is required on all carbon and low alloy materials having a specified minimum yield strength greater than 65 ksi, unless specifically exempted in Fig. UCS-66 (see UCS-66(f)). Impact testing is now required in base metal and welded joints for any thiclmess/design metal temperature below the applicable impact test exemption curve, except for certain exemptions, and adjustments to the permissible design metal temperature (based on successful past experience and fracture mechanics considerations), as listed below: • UG-20(f) exempts Pl Group 1 and 2 materials from impact testing at temperatures down to -20°F (and lower when due to seasonal atmospheric temperatures) providing they do not . exceed %" thickness for Curve A materials and l" thickness for Curve B, C and D materials and provided the vessel is hydrostatically tested per UG-99. • UCS-66(b) and Fig. UCS-66.1 permit a further basis for reduction of the minimum design metal temperature for stationary vessels (for design metal temperatures -55°F and warmer) based on ratio of the design stress in tension to the allowable tensile stress. Furthermore, impact WRC Bulletin 435 • • • • testing is not required when the m1mmum design metal temperature is colder than -55°F but no colder than -155°F and the coincident stress ratio as defined in Fig. UCS-66.1 is less than 0.4. Table UG-84.4 permits an increase in the impact test temperature due to slow loading on vessels as compared to the dynamic loading on the test specimen during the Charpy test. The permissible temperature differentials in Table UG-84.4 are conservative. UCS-68(c) permits a 30°F reduction in impact test exemption temperatures to the minimum permissible temperature from Figure UCS-66 for ASME P-No. 1 materials if postweld heat treating is performed when not otherwise required by the Code rules. No impact testing is required for ASME/ANSI Bl6.5 or ASME B16.47 ferritic steel flanges used at design metal temperatures no colder than -20°F. No impact testing is required for UCS materials less than 0.10" thick or for SA-194 and SA-540 nuts at design metal temperatures of -55°F and above. It is also important to note that the Code (UG-20(a)) now requires the determination of the minimum design metal temperature to be based on the lowest temperature to be expected in service including operational upsets, autorefrigeration, atmospheric temperature (except for lower seasonal atmospheric temperatures), and any other sources of cooling. Additional requirements have been included in UG-99 and UG-100 to preclude brittle fracture during overload testing. UG-99(h) recommends that the minimum metal temperature be at least 30°F above the minimum design metal temperature during the hydrostatic test. UG-lOO(c) requires the minimum metal temperature during the pneumatic test to be at least 30°F above the minimum design melal temperature. The 1987 revisions to the Code toughness rules have significantly reduced the risk of brittle fracture and improved the safety of Code pressure vessels. E. Review of PVRC Burst Test Results Ductile rupture or the bursting mode of failure is the most fundamental of the various failure modes considered in the Code design rules. The design margins of 4 on tensile strength and 1.5 on yield strength (at temperatures below the creep range) have been established to guard against ductile rupture in the temperature region in which ductile rupture, rather than one of the other failure modes, is the governing criterion. There was considerable interest in 1960's and early 1970's to re-evaluate the Code limits on tensile stresses to see if a new criteria is needed to provide a margin of safety against bursting which is consistent over the full range of the available materials and takes into consideration strain concentrations which occur at discontinuities, such as nozzles. Because of this the PVRC Subcommittee on Effective Utilization of Yield Strength sponsored several projects to provide a better understanding of the behavior of pressure vessels in the bursting mode of failure and to evaluate the accuracy of the modified Svensson equation on smooth cylinders with moderate strain concentrations. The project at University of Kansas included burst tests on test vessels of three types of materials with different strain hardening components. The materials were SA-240 Type 304 stainless steel, SA-516 Grade 70 carbon steel and SA-517 Grade F quenched and tempered high strength low alloy steel. These steels have strain hardening exponents of 0.585, 0.189, and 0.085, respectively. Burst tests were performed on test vessels with nozzles and with sharp notches. The earlier work by B. F. Langer11 and the burst tests at University of Kansas 12 led to the following conclusions: 1. Design of Section VIII vessels is essentially based on the bursting mode of failure instead of yielding. (The yield strength controls the design only if the yield strength/tensile strength ratio is less than 0.5 for Division 2 vessels, and 0.375 for Division 1 vessels). 2. The bursting pressure of cylindrical and spherical shells can be predicted with reasonable accuracy if consideration is given to the strain hardening properties of the material. 3. The modified Svensson equation gives a reasonably accurate prediction of burst pressures in thin wall pressure vessels despite the presence of moderate strain concentrations, such as nozzles and end closures. Moderate strain concentrations do not significantly reduce burst pressure. 4. For vessels with sharp notches, the test results showed that a direct correlation between reduction in wall thiclmess (due to a sharp notch) and reduction in burst pressure exists at least until the notch depth of 25% of the wall thickness for the steels tested. Additional tests on a higher strength steel HY-140 indicated that the linear reduction theory for that steel is applicable}eyond a notch depth of 40% of the vessel wall thickness. 5. For some vessels, particularly those from steel with a high value of strain hardening coefficient (n), and which have been designed under ASME rules, the true margin of safety on bursting mode of failure can be as low as 2.4. Decreasing the design margin from 4 to 3.5 on the ultimate tensile strength would not have a significant effect on the true margin of safety. Some European pressure vessel codes use lower margins on the tensile strength than ASME. The Evaluation of Design Margins Section VIII 13 British BS 5500 uses a factor of 1.5 on yield strength and 2.35 on tensile strength for carbon and low alloy steel vessels. The German AD Merkblatt code uses 1.5 on yield strength at temperatures below the creep range. A more detailed comparison with international codes will be performed in Phase 2. 3. Paragraph UG-23(d) F. Review of Significant Code Paragraphs 4. Paragraph UG-24 The following is a brief review of those Code paragraphs which use Code stress allowables or are affected by a change in stress allowables. The purpose of this write-up is to outline the logic used to assure that a safety concern does not arise as a result of increasing the stress allowables. Those areas which may be of any concern will be listed in part (H) of this report. 1. Paragraph UG-20(f) This paragraph exempts Pl Group 1 and Group 2 materials from impact testing down to and including -20°F (and lower when due to seasonal atmospheric temperatures) providing they do not exceed W' for Curve A materials and l" for Curve B, C and D materials and providing the vessel is hydrostatically tested per UG-99. Decreasing the design margin could invalidate the experience base used for justifying this exemption. This paragraph requires that general primary membrane shall not exceed 1.2 times the maximum allowable stress for load combinations including wind or earthquake. This will allow such stresses to go as high as 1.2/3.5 of the specified ultimate strength. This is not of any concern. This paragraph specifies casting quality factors to be applied to the allowable stress values for some cast materials. These factors are believed to be based on judgment and experience. If the stress allowables are increased these factors may have to be revised. However, it should be pointed out that casting quality has improved significantly over the years. 5. Paragraph UG-27(c)(l) The formula in this paragraph is valid only when P does not exceed 0.385 SE. With a higher value of "S," the coefficient "0.385" may have to be revised. Code Committee to look into. The same comment applies to paragraphs UG-27(c)(2) and UG-27(d). 6. Paragraph UG-28(c)(2) The formula in step 3 for determining allowable external pressure of thick cylindrical shells uses a value of "2S." The formula should be reviewed by the Code Committee to assure that a revision is not required. 2. Paragraph UG-23(c) 7. Paragraph UG-32(e) This paragraph states that the maximum primary membrane stress plus primary bending stress shall not exceed 1112 times the allowable stress value. Since the limit of % on specified minimum yield strength will be maintained, this paragraph will still assure that primary surface stresses do not exceed yield. These stresses will be allowed to go up to 1.5/3.5 of the specified ultimate strength, but this should not be of any concern. The only question regarding this paragraph is the fact that the limit is put on "stresses" and not on "stress intensities." Also the fact that a definition of "primary" stress is not provided in Division 1 is of some concern. But these are concerns about the existing paragraph regardless of stress allowables. The other statement that is of significance in this paragraph is the following: "It is recognized that high localized discontinuity stresses may exist in vessels designed and fabricated in accordance with these rules. Insofar as practical design rules for details have been written to limit such stresses to a safe level consistent with experience." This statement reinforces the fact that Code details have been developed assuming the present stress allowables. In this report (and in more detail in phase 2) the authors will review the Code details to identify any areas of concern. If the Code stress allowables are increased, the Code Committees will also need to review these details to assure that the above statement is still valid. This paragraph requires that torispherical heads made of materials having a specified minimum tensile strength exceeding 80,000 psi be designed using a value of S equal to 20,000 psi (reduced at high temperatures). This requirement has been introduced to guard against knuckle buckling, which the Code does not explicitly address. When using higher stress allowables, buckling could start becoming of concern at a lower value of D/t. Therefore, this requirement may have to be revised by the Code Committee (The Code Committee has an item to look into this requirement of the present Code). 14 8. Paragraph UG-33(f)(l)(b) The formula in step 3 uses a value of "28." This formula should be reviewed by the Code Committee to assure that a revision is not required. 9. Paragraph UG-32(f) The value of 0.665 in this paragraph is a function of the allowable stress and may have to be revised for a higher allowable stress. higher proportion of ultimate strength for these surface stresses should not be a problem. Paragraph UG-34(c) provides rules for the minimum thickness of fiat unstayed heads. As stated by note 21 to this paragraph, these rules provide safe construction as far as stress is concerned, and this is basically due to the fact that surface stresses are not allowed to exceed the minimum specified yield strength of the material. However, the statement that "Greater thicknesses may be necessary if deflection would cause leakage" will still be true and will have to be considered by the designer. 11. Paragraph UG-36 Note 22, related to section on "OPENINGS AND REINFORCEMENTS," states that "the rules governing openings as given in this Division are based on the stress intensification created by the existence of a hole in an otherwise symmetrical section. They are based on experience with vessels designed with safety factors of 4 and 5 applied to the ultimate strength of the shell material." Any reduction in the present design factor on ultimate strength will make this statement invalid. However, it is the opinion of the authors that the rules in Division 1 on design of openings for internal and external pressure are adequately safe with a design margin of 3.5 on ultimate tensile strength. The remaining part of note 22, on external loadings, remain true regardless of design factors, since Division 1 does not provide rules for design due to these loads. 12. Paragraph UG-36(c)(3) This paragraph provides exemption from reinforcement requirements of paragraph UG-37. The rules of this paragraph on the size of exempted openings and spacing requirements of multiple openings are arbitrary. The rules for exemption in Section VIII, Division 2 and in Section III are based on analytical justifications and are different. A comparison with other codes and a justification for present Division 1 rules will be provided in the phase 2 study. 15. Paragraph UG-81 The tolerance requirements of this paragraph for formed heads are also arbitrary and conservative. There is no safety concern with reducing the design margins. 16. Paragraph UG-99 This paragraph requires hydrotesting to a minimum value calculated by UG-99(b) or (c). This value has been chosen to produce the maximum cold prestressing and blunting of notches, without producing excessive deformation. However, the Code does not specify an upper limit for hydrostatic test pressure; and paragraph UG-99(d) provides an option to the Inspector to reject the vessel if it is subjected to visible permanent distortion. The vessel deformations are a function of yield strength and not the ultimate strength. Reducing the design factor on ultimate strength will allow stresses which are a higher fraction of yield strength for those materials whose allowable stresses are controlled by ultimate strength. However, the present limits on yield strength still have to be met and this will probably guard against hydrotest resulting in unacceptable deformations. Therefore, it is believed that the factor of 1.5 on MAWP for hydrotest can remain unchanged. It is recommended that the Code add limits for membrane stresses during hydrotest. For carbon and low alloy steels, a limit of 90% of the yield strength at test temperature, on general primary membrane stresses has been successfully used and is recommended. This will be consistent with the Division 2 limit on general membrane stresses. However, since bending stresses are not normally calculated for a Div. 1 vessel, no limit on surface stress may be imposed. Somewhat higher allowable test stresses should be acceptable for austenitic stainless steels and nonferrous materials, due to their lower yield strength-to-ultimate strength ratio, better ductility, and their strain hardening characteristics (provided permanent distortion is not a problem). 17. Paragraph UG-101 This paragraph provides for the use of standard flanges and pipe fittings without performing design calculations. Experience has shown that, in general, these standard flanges and fittings will not meet the Code design requirements using the present allowable stresses. Allowing higher stress allowables will probably result in Code designs being closer to standard designs which have been shown to be safe by service experience. This paragraph allows establishing the MAWP of the vessel or vessel parts using a test procedure. Of the various test procedures, the procedure of UGlOl(m) on burst test needs a review. The present rules basically call for a design factor of 5 on burst pressure. If design rules are based on a reduced factor of 3.~n ultimate strength, the factor of 5 on burst test may be excessive. It is recommended that the Code Committee review this factor to evaluate the merits ofreducing the value of 5. 14. Paragraph UG-80 18. Paragraph UF-27(b) The out-of-roundness tolerances permitted by this paragraph are arbitrary and generally recognized as conservative. Even though increased internal pressure allowable will result in higher bending stress due to out-of-roundness, these stresses are selflimiting and there is no safety concern. The rules of this paragraph allow a reduction in design pressure if out-of-roundness exceeds 1%, but not exceeding 3%. No reduction in design pressure is required if calculated bending stresses are less than 114 of design stress value, and the reduction in pressure beyond this value is proportional to the 13. Paragraph UG-44 10. Paragraph UG-34 This paragraph on design of unstayed flat heads and covers provides values for factor "C." These values of"C" include a 0.667 factor which effectively increases the allowable stress to l.5S. This will continue to limit the surface stress to specified minimum yield strength. The fact that a lower design margin on ultimate strength will allow a WRC Bulletin 435 Evaluation of Design Margins Section VIII 15 ratio of bending stress to the design stress value. If higher design stress values are allowed, proportionally higher bending stresses will be allowed. However, the bending stresses due to internal pressure are self-limiting and will not affect the safety of the vessel. 19. Paragraph UCS-6 No specific changes are needed since the materials are subject to the toughness requirements ofUCS-66 (see comments on UCS-66(b) 3 ). 20. Paragraph UCS-66 These rules for assuring adequate toughness are based on certain assumptions regarding the tensile stresses and flaw sizes present at the minimum design metal temperature of the material. These toughness rules have been discussed elsewhere in this report. A more detailed review of these requirements, in light of the proposed reduction on design margin, will be performed in Phase 2. However, it is not expected that any significant revisions to this paragraph will be needed. The Code Committee has adopted these rules, with the same exemption curves, for Div. 2 which has even a lower design margin than that proposed for Div. 1. (Div. 2 has more restrictive NDE requirements). A comparison of design, fabrication, NDE, and material requirements relating to toughness with other codes will be performed in Phase2. 21. Paragraph UCS-66(b)(3) This paragraph allows exemption from impact testing, for certain cases, if the ratio of design stress to allowable stress is less than or equal to 0.40. It is generally believed that brittle fracture is of no concern if the tensile membrane stresses do not exceed 10% of the ultimate tensile strength. The factor of 0.40 was based on the assumption that the allowable stresses did not exceed 0.25 of tensile strength (0.40 x 0.25 = 0.10). With a reduction in design margin on ultimate strength, this factor should be revised. 22. Paragraph UCS-68(a) See above comment on Paragraph UCS-66(b) 3 • 23. Paragraphs 1-2 and 1-3 of Appendix 1 These paragraphs provide design formulas for thick shells. These formulas are applicable when design pressure exceeds a fraction of allowable design stress. With a higher allowable stress the value of these fractions may have to be revised. It is recommended that the Code Committee review. 26. Paragraph 1-8(e) See above comment regarding Paragraph 1-5(g). 27. Appendix 2 It is generally recognized that the rules of this Appendix are conservative. Standard flanges usually do not meet the stress allowables of this Appendix and yet they have performed safely. Therefore a reduction in design margin will not be of any. concern. (The Code Committee is developing totally different set of design rules to replace Appendix 2 rules for flange design.) 28. Appendix 27 Paragraph 27-2 allows out-of-roundness values exceeding those of the Code, for glass lined vessels. reduction in design pressure, similar to that o Paragraph UF-27(b) is required. See above commen on UF-27(b). The rule of Paragraph 27-5 allowin SA-285 Gr.C material to be assigned to Curve should also be reviewed. It is suggested that i addition to the 0.18% maximum carbon restriction:, the SA-285 Gr.C should also be fully killed to justif its use on Curve B. 29. Code Case 2176 This Code Case allows exemption from impac testing in certain cases, provided the ratio of desi stress to allowable stress is less than 0.40. See abov comment on Paragraph UCS-66(b). 3 30. Code Case 2168 24. Paragraph 104 Note 1 of this paragraph should be reviewed by the Code Committee. (There is an item for this review on SGD/VIII agenda). 25. Paragraph 1-5 This paragraph provides rules for the design of conical reducer sections and conical heads. These 16 rules are generally conservative and reducing the design factor on ultimate strength to 3.5 will not result in any concern, except possibly for the provisions of subparagraph 1-5(g). This subparagraph, for half apex angles greater than 30°, call for special analysis. The limit on local hoop membrane stress is 1Yz times that of general primary membrane stress which is consistent with Div. 2 analysis. However, the allowable for surface stress at discontinuity is 4 times the basic allowable, whereas for Div. 2 analysis, this factor is 3. This was justifiable in light of present design factors. If the design factor on ultimate tensile strength is reduced to 3.5, it is recommended that 4SE in 1-5(g) be reduced to SSE. This will be consistent with other Code sections. (There is also some concern that some of the bending stresses at cone-cylinder junctions are primary.) Code Case 2150, which provides analytical results for such cone-cylinder junctions, calls for a limit of 3S, implicitly accounting for the joint efficiency. This Code Case should be reviewed to assure consistency with the Code allowables. This Code Case provides an alternative method t area replacement rules, based on recommendation of WRC Bulletin 335. These rules need to be re viewed to assure that an increase in stress allowabl will not invalidate the assumptions and test result used in WRC Bulletin 335 to come up with th recommendations therein. It is recommended tha WRC Bulletin 435 this be investigated in detail as a part of Phase 2 study. 31. Material Code Cases There are numerous Code Cases allowing certain materials which are not in Section II, Part D. These Code Cases provide stress allowable values generally based on present Code criteria. If the basis for Code stress allowables is revised, all such Code Cases will have to be reviewed to assure consistency. G. Conclusions There have been extensive rev1s1ons to Section VIII Code rules since the present design margins were established in 1940's. This has led to general improvement in the safety and reliability of Section VIII, Division 1 pressure vessels. No efforts, however, have been made so far to try to justify a further reduction in the design margins taking credit for the revisions and improvements in the Code rules. It is the conclusion of the authors of this report that a reduction in the present design margins from 4 to about 3.5 at temperatures below the creep range would be justified based on the improvements in the Code rules and excellent past experience with vessels built to the Code rules. A number of recommendations are made for additional work to further assure that a reduction in design margin will not result in any inconsistent rules. Section H lists certain areas suggested for Code Committee review. These will be supplemented and finalized in Phase 2. H. Suggestions for Revisions to Code Requirements The following is a list of the Code requirements which may have been based on the assumption that stresses are limited to those allowed by the preseD;t Code. If the stress allowables are increased, thes'e requirements may have to be revised by the appropriate Code Committee to assure that no revisions to the Code rules are needed. The authors have the following suggestions. 1. Paragraph UG-20(f) T~is paragraph allows exemption from impact testmg of materials which would have been required ~o be impact tested by UCS-66. This exemption was mtroduced based on extensive experience with the use of these materials and thicknesses at temperatures down to - 20°F. Increasing the allowable stress could invalidate that experience base. This paragraph should be reviewed by the Code Committee and revised as needed. It is the opinion of the authors that this exemption ma?' be continued with the design margin of 3.5 on ultimate tensile strength. This is based on the fact that similar good experience exists for relatively thin wall Ct< 1.0 in.) ASME B 31.1 piping components where the design margin is 3.0 on ultimate tensile strength. This piping code, as well as foreign code experience should also be considered in the Phase 2 effort, but a more detailed evaluation is needed. 2. Paragraph UG-24 The casting quality factors in this paragraph should be reviewed by the Code Committee, but the authors believe that there are adequate margins in these factors to justify the proposed reduction in the design margin. 3. Paragraph UG-36 Note 22, which indicates rules for design of openings are based on experience with vessels designed with safety factors of 4 or 5 applied to the ultimate tensile strength, is no longer needed and should be deleted. 4. Paragraph UG-53(i) This paragraph calls for multiplying "the appropriate stress value in tension" by the factor 1.18 for certain ligament efficiencies. This factor is to be applied to "the stress allowable in tension," and the wording should be corrected. The reason for the introduction of this factor and its particular value is not known to the authors. 5. Paragraph UG-99 The Code has no limit on stresses due to hydrotest. The criteria of "visible permanent distortion" of Paragraph UG-99(d) may not be sufficient with the lower design margin. The authors recommend limiting the general primary membrane stress to 90% of the ultimate tensile strength for all ferrous materials (except austenitic stainless steels). 6. Paragraph UG-99(h) This paragraph recommends that metal temperature during hydrotest be maintained at least 30°F above the MDMT. With a lower design ma.rgin, it becomes more essential to require a minimum metal temperature during hydrotest. The authors recommend a mandatory requirement of at least l0°F above the MDMT for thicknesses up to and including 1 inch and 30°F above MDMT for thicknesses above 1 inch. 7. Paragraph UW-12 Table UW-12 assigns joint efficiencies of 0.50 and 0.55 to singil'e and double full fillet lap joints. The design of such joints is normally controlled by fatigue failure criteria. Even though the rules of Div. 1 require consideration of cyclic loads (UG-22), a vast majority of Div. 1 vessels are considered not to be in cyclic service and no fatigue analysis is performed for such vessels. If lower design factors are allowed, it is recommended that the Code Committee review the adequacy of joint efficiency values for fillet lap joints. It may be desirable to slightly reduce these factors so that the past service experience is not violated. Evaluation of Design Margins Section VIII 17 8. General With higher stress allowables, the Code Committee will need to re-evaluate the transition from short time to long time properties (the temperatures at which the design in creep range starts governing). I. Recommendations The following -is a list of recommendations for follow-up to this study, as further justification for lower design factor on ultimate tensile strength in Section VIII, Division 1. 1. Phase 2 Study It is recommended that this study be followed by a more in-depth study. Part of the Phase 2 study could cover the following areas a. A more detailed investigation of the basis for pertinent rules in the present Code, and what effect a reduction in design margin may have on vessel safety. b. A comparison of certain parts of the Code with Section VIII, Div. 2, the BS5500, AD Merkblatt and possibly other international codes. Based on these comparisons, additional recommendations may be provided on design rules, fabrication requirements, NDE requirements, and materials requirements to justify lower design margins. c. Amore in~depth investigation of fatigue strength of vessels fabricated to the Code rules. The fatigue exemption rules of Section VIII, Div. 2 will be used as a guideline to estimate the number of pressure and temperature cycles that could safely be accommodated, without doing a fatigue analysis, for proposed design factors. d. A more in-depth study of the effects oflowering design margins on margins to brittle fracture. e. A study of additional vessel test results and, based on results of these tests, drawing conclusions on actual margins to failure of fabricated vessels. Some of the test results which will be reviewed include the two ellipsoidal heads tested for PVRC by CBI in mid 1980's, the two heads tested for PVRC by Praxair in 1994, several reduced girth vessel models tested by CBI, several fabricated models of nuclear containment vessels tested for the NRC, and the tests of vessels for the MPC Program on Fitness-ForService by Stress Engineering. f. A review of Code Case 2168 and WRC Bulletin 335 to assure safety of the alternative rules for reinforcement. 2. Code Committees As a part of any reduction in Code design factors, it is recommended that the following areas be reviewed by the appropriate Code Committees: a. The quality factor of 0.92 for structural materials. (See 1-lOO(c) of Appendix 1 of Section II, 18 Part D). With lower design margins, the Code Committee may want to discourage the use of these materials, by lowering the quality factor. b. Paragraph UG-20(£). This exemption is based on experience with present Code stress allowables. (See H.1 for authors' comments). c. Note 22 under Paragraph UG-36. This note has to be revised to reflect experience with safety factors other than 4 or 5. (See H.3 for authors' comments). d. The factor of 1.18 in Paragraph UG-53(i). Does this factor need to be revised if design margins are lowered? e. A limit on stresses due to hydrotest added to Paragraph UG-99? (See H.5 for authors' comments). f. Recommended hydrotest temperature in Paragraph UG-99(h). Should this temperature be made mandatory? (See H.6 for authors' comments). g. Joint efficiencies for fillet lap joints of Paragraph UW-12. Should these values be reduced? (See H.7 for authors' comments). h. Fabrication tolerances, weld detail requirements, repair requirements, and NDE requirements. The Code Committee should make a general review of these areas to assure that experience and judgment do not dictate any revisions iflower design margins are adopted. i. Review casting quality factors of Paragraph UG-24. With lowering design margins would these factors have to be revised? (See H.2 for authors' comments). j. Factors on "SE" in Paragraph UG-27. k. The limit of 20,000 psi in Paragraph UG-32(e). 1. The factor 0.665 in Paragraph UG-32(£). m. The safety factor of 5 in Paragraph UG-101. n. Factors in Paragraphs 1-2 and 1-3. o. Note 1 to Paragraph 1-4. p. The limit 4SE in Paragraph 1-5. q. Code Case 2150 to assure that allowable stresses are consistent with the text of the Code. r. The factor of0.40 in UCS-66(b), 3 UCS-68(a) and Code Case 2176. This factor may have to be adjusted to allow exemption at 10% of ultimate tensile strength. s. Appendix 27 and in particular Paragraphs 27-2 and 27-5. t. All material Code Cases which include material allow ables. u. Formula in step 3 ofUG-28(c). 2 v. Formula in step 3 ofUG-33(f)(l)(b). w. The proposed decrease in the design margin will not affect the allowable stress values in the creep range. The Code Committee should, how-_ ever, make a review of the transition from time independent to time dependent properties. 4. History of ASME Boiler Code, by Arthur M. Greene, Jr., published in 1952-1953 issues of Mechanical Engineering, ASME G00028. 5. ASME Boiler Construction Code Section VIII, Rules for Construction of Unfired Pressure Vessels, 1950 Ed. 6. Design Criteria of Boilers and Pressure Vessels, papers presented at the First International Conference on Pressure Vessel Technology, Sept. 99-0ct. 2, 1969, The Netherlands. ~ 7. PVP-Vol. 97-1984, International Design Criteria of Boilers and Pressure Vessels, papers presented at the Fifth International Conference on Pressure Vessel Technology, Sept. 9-14, 1984, San Francisco, CA. 8 PVP-Vol. 152-1988, Design Criteria of Boilers and Pressure Vessels, papers presented at International Conference on Pressure Vessel Technol- 0~: WRC Bulletin 101, November 1964, PVRC Interpretive Report of Pressure Vessel Research, Section 2-Materials Considerations, by J. H. Gross. 10. WRC Bulletin 158, January 1971, PVRC Interpretive Report of Pressure Vessel Research, Section 3-Fabrication and Environmental Considerations, by A. P. Bunk. 11. Design Stress Basis for Pressure Vessels, by B. F. Langer, Experimental Mechanics, January, 1971. 12. Paper No. 74-Mat-1, ASME Journal of Engineering Materials and Technology, Effect of Strain-Hardening Exponent and Strain Concentrations on the Bursting Behavior of Pressure Vessels, by C. P. Royer and S. T. Rolfe. 43~:ds,uD~~f~~~~!~f~'f~r i3"';i1~;s 7:~j'Pi'.~fs"~r~~~s;~~·in1ih~ J~~~.. b~ M.D. Bernstein. 14. Summary of Design Analysis Factors Inherent in the Established Failure Modes of the ASME Boiler and Pressure Vessel Code, ASME Subgroup Design Analysis (SCD) report, February, 1989. 15. Paper No. 88-PVP-8, New Toughness Rules in Section VIII, Division 1 oftheASME Boiler and Pressure Vessel Code, by A. Selz, paper presented at the Pressure Vessels and Piping Conference, June 19-23, 1988, Pittsburgh, PA. 16. Brittle Fracture Margins in ASME Code, Section VIII, Division 1, by S. Yukawa, Jan. 7, 1994 (Prepared for MPC Program on Fitness-for Service). 17. Structural Analysis and Design of Process Equipment, Second Ed., M. H. Jawad and J. R. Farr, 1989, John Wiley & Sons, Inc. 18. The International Journal of Pressure Vessels and Piping, Vol. 1, No. 2, 1973, Failures of Boilers and Pressure Vessels: Their Causes and Prevention, by J. F. Lancaster. 19. Journal of Pressure Vessel Technology, Vol. 110, August 1988, pp 226-233, Statistics of Pressure Vessel and Piping Failures, by S. H. Bush. 20. BWRA Bulletin (Special Report), Vol. 7, No. 6, June 1966, Brittle Fracture of a Thick Walled Pressure Vessel. 21. Journal of Engineering Materials and Technology, Vol. 102, October 1980, pp 384-387, Brittle Fracture of a Heat Exchanger Shell: A Case History, by R. F. Wagner and D. R. Mcintyre. 22. Welding and Metal Fabrication, January 1973, pp 4-12, Pressure Vessel Failure During Hydrotest, by B. Banks. 23. Materials Pe1formance, Vol. 20 (No. 9), 1981, pp 2. Failures due to poor notch toughness, fabrication defects, welded or welded repairs. 21-26, The Application of Fracture Mechanics to Refinery Failure Analysis, byW. R. Warke. 24. Brittle Fracture of a Pressure Vessel-Study Results and Recommendations, by R. D. Merrick and A. R. Ciuffreda, presented at the API 48th Midyear Refining Meeting, May 10, 1983, Los Angeles, CA. 25. Failures During Hydrotest, September 20, 1983, CBI in-house con-espondence. 26. NBSIR 86-3049, Examination of a Pressure Vessel That Ruptured at the Chicago Refinery of the Union Oil Company on July 23, 1984, by H. I. McHenry, T. R. Shives, D. T. Read, J. D. McColskey, C. H. Brady and P. T. Purtscher, National Bureau of Standards, March, 1986. 27. Pressure Vessel and Piping Design-Collected Papers, 1927-1959, ASME Publication, 1960. 28. Pressure Vessel and Piping Technology-A Decade of Progress, 1982. 29. Coke Drum Failure During Hydrostatic Testing, presented at API Operating Practices Committee, May 14, 1973, Philadelphia, PA. 30. G. G. Karcher,ASME TG-Toughness, Subcommittee VIII, February 13, 1986 (Committee Correspondence). 31. Sulfur-Some Effects on Steel Processing and Steel Properties, by G. J. Roe, P. R. Slimmon and G. F. Melloy, Proceedings, 55th National Open Hearth and Basic Oxygen Steel Conference, April 10-12, 1972, Chicago, IL. 32. High-Strength, Low-Alloy Steel-A Decade of Progress, By F. B. Pickering, Micro Alloying 75. 33. Inclusions and Mechanical Properties, by W. C. Leslie, ISS Transactions, Vol. 2, 1983-1. 34. Experimental Effort on Bursting of Constrained Discs as Related to the Effective Utilization ofYield Strength, by W. E. Cooper, E. H. Kottcamp, and G.A. Spiering,ASME 71-PVP-49. 35. Properties and Characteristics of a Quenched and Tempered Steel for Pressure Vessels, by W. D'Orville Doty, Welding Journal Research Supplement, September 1955, pp. 1-17. 36. Design of Welded Pressure Vessels Using Quenched and Tempered Steel, by Leonard P. Zick, Welding Journal Research Supplement, September 1955, pp. 18-24. 37. Suitability of Quenched and Tempered Steels for Pressure Vessel Construction, by Leon C. Bibber, Welding Journal Research Supplement, September 1955, pp. 25-40. 38. WRC Bulletin 407, December, 1995. 39. NBSAR 86-3049, Examination of a Pressure Vessel that Ruptured at the Chicago Refinery of the Union Oil Company on July 23, 1984, by H.I. McHenry, T.R. Shieves, D.T. Read, J.D. McColskey, C.H. Brady, and P.T. Purtscher. Bibliography 1. ASME Code, Section VIII, Division 1, 1992 Ed. 2. ASME Code, Section II, Part D, 1992 Ed. 3. ASME Code Cases: Boilers and Pressure Vessels, 1992 Ed_ WRC Bulletin 435 Evaluation of Design Margins Section VIII 19 Report No . 2: Evaluation of Design Margins for ASME Code Section VIII, Divisions 1 and 2-Phase 2 Studies E. Upitis and K. Mokhtarian A. Executive Summary The Phase 1 Report on Evaluation of Design Margins for ASME Code Section VIII, Division 1, May 1996, provided justification for decreasing the design margin on ultimate tensile strength from 4 to 3.5. n That report endorsed the reduction in the design margin based on the improvements in the Code rules over the last 40-50 years and recommended further review of certain other factors to assure consistency of the existing rules with the reduction in the design margin from 4 to 3.5 for Division 1. The Phase 2 Report includes the Phase 1 follow-up studies, evaluation of the design margins in ASME Section VIII, Division 2, and a comparison of the Section VIII, Division 1 and Division 2 Code requirements with those in several European pressure vessel codes and standards which use lower design margins than ASME Section VIII. These include the British Standard 5500, 5 the German Pressure Vessel Code (AD-Merkblatt Code), 6 the Netherlands Rules for Pressure Vessels (Stoomwezen Code)? and the CEN/TC54 Draft Standard for Unfired Pressure Vessels. 8 The design rules in ASME Section VIII, Division 1 do not explicitly address the fatigue mode of failure. Many pressure vessels, however, are exposed to some pressure :fluctuations. The Draft CEN Pressure Vessel Standard requires all pressure vessels to resist 500 full pressure cycles. Section C.2 of this report attempts to evaluate the effect of reduced design margins on fatigue strength of Division 1 vessels. The use of SA-516 Gr. 70 material, Division 1 allowable stresses and a stress concentration factor of 4 for reinforcing pad fillet welds, results in approximately 600 allowable pressure cycles using the Division 2 fatigue design curves. This reduces to about 500 for design margin of 3.5 and about 190 for 20 design margin of 3. The use of a design factor less than 3.5 necessitates the use of improved details or a restriction on the allowable fatigue cycles. The basic philosophy behind the area replacement rules in Division 1 is to assure that the burst pressure of the vessel is at least equal to a vessel without any openings. The recent Code Case 2168 allows less than full area replacement, but is limited to materials with allowable stresses in tension not exceeding 17 .5 ksi and has a provision that cyclic loading is not a controlling design requirement. A PVRC project will include a detailed evaluation of Code Case 2168 and will review the differences between ASME Section VIII and the European pressure vessel codes. The report by Dr. S. Yukawa on Brittle Fracture Margins in the ASME Code Section VIII, Division 1 evaluated margins on brittle fracture for 38 and 50 ksi yield strength steels based on the ratio of the available dynamic fracture toughness Km to the applied K1, using the Division 1 allowable design stresses and the maximum allowable flaw sizes in Division 1 for radiographic examination. 27 Section C.4 of this report provides additional information on brittle fracture margins for the high strength, Q & T. SA-517 steels in Part UHT of Division 1 (which are subject to lateral expansion requirements in the Charpy test), using a similar approach to that by Dr. Yukawa, with the additional consideration of residual stresses in welded joints. A total stress (design membrane stress plus welding residual stress) equal to the yield strength of the weld was used for as-welded joints. A residual stress of30% of the yield strength of the weld was added to the design stresses for the postweld heat treated (PWHT) joints. With a design factor of 4 on tensile strength (Suf4), th calculated margins on brittle fracture were about 1. for the %" material and 1.2 for the 1W' material i WRC Bulletin 435 the as-welded condition. In the PWHT condition, with a residual stress equal to 30% of the yield strength of the weld, the calculated margins on brittle fracture were 2.1 for the 1114" weld and 2.0 for the 2Y:i" weld. Reducing the design factor to 3.5 reduced the margin on brittle fracture to 1.9 for both the 1114" weld and the 2Y2" weld in the PWHT condition. The calculated margins on brittle fracture were considerably higher without considering residual stresses in welded joints. Section D of this report summarizes burst test results on several test vessels and heads. The test vessels described in Section D were all designed to Section VIII, Division 1 allowable stresses. Some were scale models and others were full size service exposed cylindrical vessels with local thin areas (LTA's). The cylindrical vessels failed at 3.0 to 4.5 times the allowable design stresses, with average being about 3.7. The safety margins for the formed heads were higher than those for the cylindrical vessels. The failure modes considered in ASME Section VIII, Division 2 are ductile rupture, brittle fracture, fatigue, incremental collapse, elastic instability and plastic instability. 25 Section E of this report includes an assessment of higher design stresses on ductile rupture and brittle fracture. It also includes. comments on the significance of yield-to-tensile strength ratios (for high strength quenched and tempered pressure vessel steels) and the Code requirements for fabrication and NDE. Ductile rupture is the most fundamental mode of failure considered in the Code rules. PVRC sponsored a project at University of Kansas in 1970's to develop a better understanding of the bursting mode of failure of pressure vessels. 19•20 The project included burst tests on test vessels fabricated from three materials with different strain hardening exponents (Type 304 stainless, SA-516 Gr. 70 and SA-517 Gr. F). One set of test vessels was tested wilth moderate strain concentrations (nozzles, fiat heads) and another set with sharp notches at various depths in the axial direction of the vessels. These tests confirmed that the modified Svensson equation (which includes the effect of strain hardening of the material) can be used to predict the burst pressure in vessels with moderate strain concentrations and that there is a direct correlation between reduction in the wall thickness and reduction in burst pressure for notch depth up to at least 25% of the wall thickness. Using the modified Svensson formula and the specified minimum tensile properties, the theoretical margins of safety for the University of Kansas test vessels (without sharp notches) designed to Section VIII, Division 2 rules, ranged from about 2.8 for the 304 stainless steel vessel (with high strain hardening exponent) to about 3.2 for the high strength, quenched and tempered steel SA-517 (with low strain hardening exponent). The actual margins were higher because the actual tensile properties exceeded the specified minimum properties. Increasing the design stress from Su/3 to Su/2.4, reduced the margin from about 3.2 to 2.6 for the SA-517 steel and from about 3.0 to 2.75 for the SA-516 Gr. 70 steel. Therefore, it does not appear that a reduction in the present ASME Section VIII, Division 2 design margin against the bursting mode of failure to a lower design margin (similar to those used in some European pressure vessel codes) would compromise the safety of pressure vessels, providing the other considerations for increased design stresses discussed in this report are met. Fracture mechanics calculations were performed to evaluate the effect of higher design stresses (Su/ 3.5 vs. Su/4 and the lesser of Su/2.4 and % and % Sy vs. Su/3) on critical flaw sizes and to determine the toughness needed to insure the same critical flaw sizes with the higher design stresses as with the present Code allowable design stresses (see Section E.2). The required fracture toughness (Krc) was evaluated for base metal and for welded joints. The required fracture toughness for the increased design stresses in base metal (with primary membrane stresses only) was evaluated with two methods, the BS PD 6493:1991 Level 2 failure assessment diagram (FAD) procedure and the ASME Section XI, Article A-3000 method. Increase in design stress from Sul 4 to Sul 3.5: The increase in the required toughness was 16-23% with the BS PD 6493 Level 2 FAD procedure and 14.515.5% with the ASME Section XI procedure for the materials and thicknesses considered in this evaluation. It is recognized, however, that there is a significant amount of successful experience with ASME Section VIII, Division 2 vessels designed to higher stresses (Su/3) and essentially the same toughness requirements as Division 1 vessels. Therefore, no increase in the Code toughness requirements is proposed in this report for increasing in design stress from Su/4 to Su/3.5. However, a further review of the toughness requirements for some of the materials in Table UHT-23 is suggested, regardless of the design stress (e.g. SA-517). Increase in design stress from Sul 3 to the lesser of Suf 2.4 and % Sy: This resulted in the following increases in the required toughness: SA-516 G~O: 11-13.5% with the PD 6493 Level 2 FAD procedure (9-10% with theASME Section XI method). SA-537 and SA-517: 28-35% with the PD 6493 Level 2 FAD procedure (26.5-27.5% with the ASME Section XI method). As-welded joints in this case were assumed the have a residual stress equal to the yield strength of the weld in the direction of the weld (assumed to be the same as the yield strength of the base metal for these calculations). Stress relieved welds were assumed to have a total stress equal to the design stress plus Evaluation of Design Margins, Section VIII 21 30% of the yield strength of the weld. The increase in toughness needed for stress relieved welded joints was somewhat smaller than for base metal, since the percent increase in the total stress was lower. The initial fl.aw sizes in as-welded and in stress relieved welded joints, however, are also smaller. The calculations in Section E.2 resulted in small critical fl.aw sizes for the high strength Q & T SA-517 steels, particularly in the as-welded condition with high residual stresses. This indicates that a further review of the Section VIII toughness and/or NDE requirements may be desirable for some of the high strength steels_ in Table UHT-23 of Division 1 and Table AQT-1 of Division 2, regardless of the allowable design stresses. A significant part of this report consists of review of the British Standard 5500, 5 German (AD-Merkblatt) Pressure Vessel Code, 6 Netherlands Rules for Pressure Vessels (Stoomwezen Code) 7 and the CEN/ TC54 Draft Standard for Unfired Pressure Vessels8 and comparison of the pertinent requirements in these codes and standards with those in ASME Section VIII, Divisions 1 and 2. A summary of each of these codes and standards is included in Appendices 1through4 of the PVRC report on design margins. 40 (Appendices 1-4 are not included in this WRC Bulletin). Section F of this report includes a comparison with ASME Section VIII, Division 1 and 2 requirements. All these European pressure vessel codes and standards have more liberal design margins than ASME Section VIII. They use a factor of 1.5 on the specified minimum yield strength at design temperature and the following factors on the specified minimum tensile strength at room temperature: • 2.35 in BS 5500 for Construction Categories 1 and2, • 2.27 in Stoomwezen Code (for materials with elongation> 10%), • 2.4 in Draft CEN/TC54 Standard The AD-Merkblatt Code does not give a factor on ultimate tensile strength for materials below the creep range, except for materials without a guaranteed yield strength or elastic limit (such as gray cast iron and copper and its alloys). The basic design formulas for shells are essentially the same in the different codes and standards. However, the formulas for design of components and for areas of discontinuity vary a great deal. In general, the ASME Section VIII Codes (especially Division 1) are more conservative than those of the European Codes studied. In most European codes the rules for design of cone-cylinder junctions are based on limit analysis or other analytical methods. The rules for nozzle reinforcement vary a great deal also. The European codes and standards use more analytical methods, such as limit analysis. The fatigue curves in Section VIII, Division 2 are 22 based on smooth specimen tests, and the fatigue analysis rules need updating. The European rules are based on fatigue tests on actual weldments and are comprehensive. The Draft CEN Standard and the Stoomwezen Code classify vessels in risk or hazard categories, depending on the toxicity of the fluids in the vessel, design pressure and volume. Vessels in the high risk categories require 100% NDE of all welds. Most of the European pressure vessel codes and standards include special requirements for materials. Some examples are: • In Holland, only materials specifications reviewed and approved by the Competent Body can be used for pressure vessels. In Germany, the suitability of filler metals and consumables for certain groups of pressure vessels must be established by an expert's report. • CEN requires European technical approval of material manufacturers and a review and approval of the material manufacturer's quality system by the Authorized Inspection Agency. • Generally, more testing is required by the European codes. The AD-Merkblatt Code requires testing both ends of plates if they exceed a certain length. The AD-Merkblatt and Stoomwezen Codes require testing of plates after forming. • The Stoomwezen Code requires all non-alloyed and low alloy steel materials to be supplied in the normalized, normalized and tempered, or in the quenched and tempered condition. • The AD-Merkblatt and Stoomwezen Codes require all materials to be impact tested • Several European pressure vessel codes require the type and extent of repair welding to be indicated in the test certificate. • CEN requires review and approval of the material manufacturer's quality system by the Authorized Inspection Agency. The toughness requirements in the European pres-. sure vessel codes and standards differ from those in ASME Section VIII. The toughness requirements in BS 5500 are based on Charpy V-notch correlations with wide plate test results. These codes provide separate impact test exemption curves for as-welded and for PWHT construction. The impact test tempera-. ture depends on the calculated reference or assessment temperature, thickness and whether the welded assembly is stress relieved. Formulas for the refer~ ence or assessment temperature include adjustments for calculated membrane stress, construction categories (in case of BS 5500) and PWHT of vessel assemblies. Substantial benefits are given to PWHT. BS 5500 includes toughness requirements only for operating temperatures below 0°C. The required minimum average impact values are 27 J (20 ft-lbs) for steels with specified minimum UTS < 450 N/mm2 and 40 J for steels with specified minimum UTS? WRC Bulletin 435 450 N/mm 2 • BS 5500 does not require impact testing of heat affected zones (HAZ) with multi-pass welding with welding heat inputs between 1 kJ/mm (25 kJ/in.) and 5 kJ/mm (127 kJ/in.). Stoomwezen requires two types of impact testing: e Quality test at 20°C for all non-alloyed and low alloy steels. • Extra testing at the impact test temperature as determined from figures which relate assessment temperatures, thicknesses and impact test temperatures. The Draft CEN Standard lists three alternative methods for impact test requirements: Method 1 is applicable to steels with specified minimum UTS ::::; 460 N/mm 2 and maximum thickness t::::; 30 mm (l.18") for as-welded construction and t ::::; 60 mm (2.36") for PWHT construction. The impact test temperature is based on the minimum design metal temperature and an adjustment based on actual tensile stresses (% of maximum allowable). Method 2 is applicable to all thicknesses and requires base metal, weld metal and HAZ to meet the required impact test values at the impact test temperature obtained from the applicable figures/ exemption curves, based on the reference temperature and thickness. Separate figures/exemption curves are provided for as-welded assemblies and for PWHT assemblies. As-welded assemblies generally must be impact tested at lower temperatures than PWHT'd assemblies. Method 3 is a fracture mechanics analysis and may be used for materials not included in this CEN standard and cases not covered by Methods 1 and 2. Some of the European codes and standards have more severe toughness requirements than ASME Section VIII Codes, particularly for thicker plates in the as-welded condition. ASME Section VIII Cod~, however, have special toughness requirements for high strength steels (minimum lateral expansion and drop weight tests for some of the high strength steels in Part UHT of Division 1 and Article F-6 of Division 2). BS 5500 and the Stoomwezen code generally do not require impact testing of heat affected zones of procedure qualification and production test plates, whereas this is generally required for vessels built to ASME Section VIII (if the base metal must be impact tested). The NDE requirements in the European codes depend on material, thickness, type of service, design stress and other considerations. For Construction Category 1 vessels, BS 5500 requires 100% RT or UT of all shell butt welds, regardless of thiclmess 100% RT or UT of nozzle attachment welds when th~ thinnest part welded exceeds the specified thickness limit for a particular materials group, and 100% MT or PT for all attachment welds. A lesser amount of NDE is required for Construction Category 2 vessels, and only visual examination is required for Construction Category 3 vessels. The NDE requirements in BS 5500 appear to be comparable to those in Section VIII, Division 2. The type and extent of the NDE in the ADMerkblatt Code depends on the type of material (material group), thickness and design stress. Generally 100% RT or UT is required for all longitudinal shell butt welds and weld intersections and 100% or a lesser amount (e.g. 25% or 10%) for circumferential butt welds. More examination is required for higher strength materials and for alloyed steels (e.g. 100% RT or UT of all shell butt welds and weld intersections). The amount of RT or UT examination, however, can be reduced if the results of previous NDE on similar vessels reveal no serious deficiencies. (This is not an option for ASME Section VIII vessels). AD-Merkblatt requires MT or PT examination of all nozzle and other attachment welds. The extent of NDE requirements in the Stoomwezen Code depends on the hazard category, material group and thickness of the pressure part. Generally, more examination is required for higher hazard categories and thicker material. Full (100%) examination is required for all Hazard Category 3 vessels and for vessels of Category C and Category CP materials, which require special care in processing or are crack sensitive materials. The NDE requirements in the Draft CEN Standard depend on the Risk Category, Testing Category of welded joints, group of steel, type of joint and maximum thickness. Longitudinal and circumferential butt welds require 100% RT or UT for some Testing Categories and a lesser amount for others. Butt welds which are RT or UT examined also require a certain amount of MT or PT examination. A unique feature in the CEN standard is that in many cases the required amount of RT (or UT) and MT (or PT) can be reduced based on satisfactory prior experience (similar to AD-Merkblatt). In general, ASME Section VIII, Division 2 appears to have the most comprehensive and well defined NDE requirements. The NDE requirements in the European pressure vessel codes and standards generally are more extensive than those in ASME Section VIII, Division 1. The out-of-roundness tolerances for shells in the European pressure vessel codes and standards appear to be '!timilar to those for ASME Section VIII vessels. However, some of the European codes and standards have additional tolerance requirements for peaking at welded seams, local fl.at areas, vessel straightness and workmanship requirements which are not included in Section VIII, Divisions 1 and 2. The European pressure vessel codes and standards generally have lower thiclmess limits for non-PWHT construction (for C and C-Mn steels) than ASME Section VIII. Also, the PWHT temperatures and hold times generally are lower (except for BS 5500, which has similar requirements to those in Evaluation of Design Margins, Section VIII 23 ASME Section VIII). BS 5500 and the Draft CEN Standard require PWHT of C and C-Mn steels over 35 mm (l.38"), AD-Merkblatt Code over 30 mm (l.18") and Stoomwezen Code over 32 mm (l.26"). Most of these codes and standards also permit increasing the maximum thickness limit without PWHT with additional requirements (additional notch toughness· or fracture mechanics analysis, additional NDE, etc.). The European pressure vessel codes and standards (particularlytheAD-Merkblatt and the Stoomwezen Codes and the Draft CEN Standard) generally require more third party involvement than ASME Section VIII, Divisions 1 and 2. Depending on the particular Code or Standard, this may include verification of design reports and materials, approval of welders, welding operators and welding procedure qualifications, witnessing or supervision of NDE, review of NDE reports and heat treat records, conducting or witnessing of pressure tests, certification of vessels, review and certification of quality management systems. The Draft CEN Standard requires the Notified Body (Inspection Authority) to verify that all parts of the vessel have been designed in accordance with the CEN Pressure Vessel Standard and that the vessel has been constructed and tested in accordance with this standard. The extent and level of the participation by the Notified Body depends on the vessel risk category, testing category and the module of conformity assessment. There is a reduced involvement by the Notified Body for a conformity assessment module involving a certified quality system. Section G of this report includes recommendations to the ASME Code Committees for revisions to Section VIII, Division 1 to justify a further reduction in the design margin below 3.5 on the ultimate tensile strength. The authors feel that this margin can be reduced down to approximately 3.0, depending on the evaluation and improvements adopted by the appropriate ASME Code Committees. No reduction has been proposed in this report for the design margin based on yield strength nor for the present design margins based on creep and rupture failures. Section G also includes recommendations for revisions to Section VIII, Division 2 to justify reduction of its design margin below 3 on the ultimate tensile strength. The authors feel that a design margin in the range of 2.3 to 2.4 may be justifiable for the new design margins providing the recommendations in this section of the report are adopted by the Code Committee. The actual design margins for both Division 1 and Division 2, however, are to be selected by the appropriate Code Committees based on their judgment and collective experience. The conclusion of this report is that the design margin on ultimate strength can be smaller than the current values for both Division 1 and Division 2 of Section VIII of the ASME Code. However, several areas in these codes need to be reviewed and im24 proved when the design margins are reduced below 3.5 on the specified minimum UTS for Division 1 and below 3 for Division 2 (see Section G). To provide further information for the Code Committees to make such decisions, additional studies by PVRC or other organizations are recommended in Section H of this report. Based on the Phase 2 study (including a comparison with the European pressure vessel codes and standards), specific recommendations are provided for adopting lower design margins for both Division 1 and Division 2. Recommendations are also provided for additional studies to further improve the Section VIII Codes and justify the proposed reduction in design margins. B. Introduction The present ASME Code, Section VIII, Division 11 design margins have existed since 1942/1943 when the design factors on the ultimate tensile strength were reduced from 5 to 4. The successful experience with Section VIII, Division 1, improved materials, more sophisticated design rules and improved vessel fabrication practices prompted a review of the rules in this Code. The Phase 1 report of this project concluded that the present design margin in Section VIII, Division 1 can be reduced to 3.5 without affecting the safety of vessels. 11 The Phase 1 report also recommended a review of certain paragraphs in the Code to assure consistency with the proposed design margin of 3.5. (At the time of this writing, the reduction of the design factor to 3.5 and the proposed reviews are underway). The main purpose of the Phase 2 report is to review the basis of the present design margins and to provide justification for future reduction of the design margins in ASME Section VIII, Division 2. This report also includes discussion of several additional issues related to the design margins for Division 1 vessels and recommendations for a further reduction of the design margins in Division 1. Section VIII, Division 2 (Alternative Rules) was issued in the 1960's and permits lower design margins than Division 1 (a factor of 3 on tensile strength and 1.5 on yield strength) and includes more comprehensive requirements than Division 1. 2 This report reviews the failure modes considered in Division 2 and discusses the significant factors which affect the design margins. This includes ductile rupture, brittle fracture, yield-to-tensile strength ratios, fabrication and NDE requirements. Several European pressure vessel codes use lower design margins than ASME Section VIII. This includes British Standard 5500, 5 the German AD-. Merkblatt Code, 6 the Dutch Stoomwezen Code7 and, the CEN Draft Standard for Unfired Pressure Ves-. sels.s These codes and standards (except CEN) have been in existence for some time and generally have a good performance record. This report reviews the pertinent requirements in the European pressure vessel codes and standards and compares them to the requirements in the ASME Boiler & Pressure. Vessel Code, including the design margins, design rules, design details, materials requirements, toughc ness requirements, fabrication requirements, overload testing, and quality control requirements and other requirements important for the safety of the vessels. WRC Bulletin 435 C. Design Margins for Section VIII, Division 1 (Phase 1 Follow-up) 1. Design Criteria of Section VIII, Division 1 The rules of ASME Section VIII-1, using design by formula approach, are based partly on analytical derivations, partly on test results, and to a great extent on experience. The rules have evolved over many decades and improved along the way with many new requirements. One of the most significant improvements has been the introduction of impact testing requirements in the 1987 Addenda. One of the weak points of this code is the fact that there still are no fatigue analysis or fatigue exemption rules introduced. The design rules and details have been developed to limit the magnitude oflocal discontinuity stresses and implicitly protect against fatigue failure for a nominal number of cycles. But that number has never been quantified. The primary mode of failure considered is ductile rupture. Other modes of failure have been addressed (see Phase 1 report) but not on a consistent and totally analytical basis. The approach to local stresses, which may result in ratcheting and fatigue, is summarized in paragraph UG-23 (c) which in part states that "It is recognized that high localized discontinuity stresses may exist in vessels designed and fabricated in accordance to these rules. Insofar as practical, design rules for details have been written to limit such stresses to a safe level consistent ;;ith experience." Thi~ paragraph al~o requir<:f.S that ... loads shall not mduce a combmed maXimum in primary membrane stress plus primary bending stress across the thickness which exceed 1Yz times the maximum allowable stress value in tension." However, there is no detail specified for the application of this rule. Some of the rules (such as design of flat cover) incorporate this rule, but usually the local stresses and bending stresses are not addressed. Some of the more recent rules consider local stresses and assure shakedown. Many of this code's design rules are in the process of being revised or alternatives are being provided for the older rules. Some others are in need of updating. The flange design rules are based on the old Taylor Forge approach, do not consider intersection of bolts and flanges, do not address leakage, and use out-ofdate gasket factors. A new set of rules is being developed which will have a more analytical ap~roach, will have leak tightness factor and will mcorporate recent gasket test data. Any reduction in design margins will have to be reflected in these proposed rules. The rules for design of formed heads also are based on some old test results, supplemented by experience base. These rules provide a non-uniform margin of safety over the range of parameters and are conservative in may cases. A more analytically based approach is being considered by the ASME Code Committee. This new approach is based on a non-linear elastic plastic analysis that accounts for stiffening effects of internal pressure on the shape of the knuckle. This proposed method assures shake clown to elastic action and implicitly allows for approximately 500 cycles of pressure application. The proposed rules at this time assume stress limits of the present code. The rules for design of cone cylinder junctions have evolved over the years and many recent improvements have been made to them to assure that collapse and ratcheting are precluded. A Code Case, based on shell analysis, has been published to provide design factors for a certain range of parameters.4 The proposed buckling rules, based on the WRC Bulletin 406, include rules for design of cone cylinder junctions under external pressure. 16 These rules, to preclude buckling and plastic collapse of the junction, are based on analytical results, modified by knockdown factors based on test results. As a result of this evolutionary process, the rules are scattered in different areas, and do not provide a consistent and analytically defensible set of rules. These rules could be improved by a totally new look in the light of recent analytical capabilities and available test data. The Code rules for external pressure and buckling design also do not cover all geometries and loadings of interest. The basis of external pressure design is summarized in ASME Section II, Part D, Appendix 3. These rules are based on theory of stability, modified by test results and experience. The "knockdown factors" (buckling strength reduction factors) applicable to theoretical solutions, to account for fabrication imperfections, have a very significant effect on resulting rules. Since the rules were put into the Code several decades ago, a great deal of test data has become available. The present rules, mostly conservative, have a factor of safety which varies significantly for various geometries and loadings. A proposed Code Case is being considered by the ASME Cocft!'Committee which will provide an alternative to basic Code rules and is based on recent data. 16 It is the intent that after a successful trial period these alternative rules replace the present rules. (Code Case 2286, approved July 17, 1998). In general, while Division 1 can remain a "design by formula" code, the design rules can be improved by taking advantage of available analytical methods. Design rules can still be in the form of simple formulas or graphs, but would have an analytical underpinning. The analytical results, where needed, may be adjusted by factors based on test data and Evaluation of Design Margins, Section VIII 25 experience. This is the approach that has been taken by most European codes and standards. With more confidence in the accuracy and uniformity of the rules, the design margins may safely be reduced. 2. Effect of Design Margins on Fatigue Strength Section VIII, Division 1 requires that cyclic loadings be considered; but does not provide any rules for analysis. There are no guidelines on what constitutes cyclic service. Every pressure vessel is subject to some cyclic loading, but usually a fatigue analysis is performed only when the number of cycles of pressure or temperature is more than a nominal number. There is no uniformity in the Code rules when fatigue analysis is required, and this is one of the weaknesses of Section VIII-1. When a fatigue analysis is required, normally the rules of Section VIII-2 are used. To determine if a vessel may be exempt from detailed fatigue analysis, sometimes the fatigue exemption rules of paragraph AD-160 of VIII-2 are employed. Par. AD-160.1 provides for exemptions based on operating experience. This is probably the more common basis used for not requiring fatigue analysis of Division 1 vessels. However this is not a very justifiable approach, since the weld details, local discontinuities, material properties, operating conditions and many other factors may be different from one vesselto another. The method of AD-160.2 is a simplified and conservative method of assuring that a vessel designed to the rules ofVIII-2 can withstand a certain number of cycles. The rules of this paragraph are applicable to integral parts of Division 2 vessels. This means that this method is not applicable to such details as pad reinforced nozzles and should not be used directly for a Division 1 vessel. These exemption rules assume a stress concentration factor of 2.0 and allowable stresses of Division 2. Many details in Division 1 vessels produce stress concentration (and fatigue strength reduction) factors exceeding 2.0. Paragraph AD-160.2 provides two different conditions for exemption from fatigue. Condition A is satisfied if the total number of cycles of various cyclic loads do not exceed 1000. These cycles consist of full-range pressure cycles, operating pressure cycles in which range of pressure variation exceeds 20% of the design pressure (it is assumed that stresses due to cycles with a lower range fall below the material's endurance limit), and effective thermal cycles. The equivalent number of effective thermal cycles is calculated by a factor assigned to various values of temperature differentials. For temperature differentials between adjacent points (for definition see AD-160.2) up to 50°F, the factor is O and for very large temperature differentials (in excess of 450°F) the factor is 20. The number of temperature cycles for components involving welds between materials having different coefficients of thermal expansion is also to be included to satisfy condition A. Condition B is a more accurate 26 method than condition A. This condition is satisfied if a number of requirements are met. The design number of full range pressure cycles may not exceed the allowable number of cycles which correspond to the primary plus secondary stress intensity range allowable (3 Sm) multiplied by an implicit stress concentration factor of 2.0. Another requirement is that the range of pressure cycles during normal operation does not exceed certain value. This assures that the maximum peak stress due to these operating :fluctuations does not exceed the stress range allowed for 106 cycles. This was based on the assumption that the stress amplitude at 106 cycles constitutes the endurance limit. (It is now known that this is not true and that the allowable stress amplitude declines with increasing number of cycles far beyond 106 in certain environments). Another requirement is a limit on temperature difference between any two adjacent points. There is also a limit on the range of temperature difference between any two adjacent points, depending on the specified number of significant temperature difference :fluctuations. Another requirement is a limit on the range of mechanical loads. All these requirements are again based on assuming a surface stress of 3Sm multiplied by a stress concentration factor of 2. The rules of AD-160.3 allow exemptions from fatigue analysis for non-integral parts (such as reinforcement pads). These rules require a more detailed calculation of stresses. This paragraph limits the total number of cycles of various operating conditions to 400. This is due to the fact that a higher stress concentration factor is assumed. The cycles with a range exceeding 15% of the full design pres~ sure range are assumed to contribute to fatigue damage. The design rules of VIII-1 do not explicitly 01' implicitly consider the fatigue mode of failure. Some of the details allowed by this code, such as nozzle attachment details, head or flange attachment de tails, or fillet welds for other attachments, produce stress concentration factors up to 4.0. (Although the maximum value of fatigue strength reduction factor is 5.0 for a crack-like defect, VIII-2 suggests a value 4.0 for the notch due to a fillet weld.) For a vessel fabricated from commonly used SA-516 Gr.70 mate rial, allowable stress is Yi of ultimate strength o 17 .5 ksi. Assuming that at discontinuities the maxi mum surface stress is 3S and the SCF is 4.0, wil result in a maximum stress range of 210 ksi. This i equivalent to a stress amplitude of 105 ksi, for whic the number ofallowable cycles (from figure 5-110.1 VIII-2) is approximately 600. For the proposed d sign factor on ultimate strength of 3.5, the resultin stress amplitude would be 120 ksi and the allowabl number of cycles would be approximately 500. For a assumed design factor of 3, the allowable number cycles would be approximately 190. For lower desi factors, the allowable based on yield strength wi start controlling and will result in allowable numbe WRC Bulletin 435 of cycles of about 170. This demonstrates that by increasing the allowable stresses, the number of design cycles decline significantly. If a factor on ultimate strength of less than 3.5 is used, the allowable details will have to be more restrictive or definite limits need to be put on the number of cycles exempt from detailed fatigue analysis. If the details are such that the resulting maximum stress concentration factor is 3.0, the fatigue strength will be the same as that of present rules if the design factor on ultimate is reduced from 4 to 3. This of course is a very simplistic analysis and does not address various types of cycles. It provides an insight into the magnitude of cycles that Division 1 vessels would be safe to be subjected to. The fatigue rules and S-N curves of Division 2 are about 30 years old. Agreat deal of additional technology and data are available today. If the design margins are reduced, it is recommended that a review ofVIII-2 fatigue analysis rules be made and, where necessary, these rules be updated. The fatigue exemption rules will also have to be revised to reflect higher allowables, resulting in a lower number of cycles at which exemption is allowed. With any increase in stress allowables for Division 1 vessels, some guidelines will need to be provided for exemption from fatigue. Some suggestions on operational cycle accounting and in-service inspection may also have to be added, until post construction codes are available to address these issues. 3. Review of Code Case 2168 and Nozzle Reinforcement Requirements All codes and standards reviewed contain rules for reinforcement of openings larger than a certain size. ASME codes almost always require full replacement of the area needed to resist pressure, within certain limits. Most European codes base their rules on more analytical methods such as limit analysis. For the comparisons of design rules in this report, the basic ASME method of full area replacement was used? However, recently a Code Case has been introduced, which allows less than full area replacement. This Code Case 2168 is based on studies of and recommendations from WRC Bulletin 335. 15 The Code Case is limited to carbon and low alloy steels with allowable stresses in tension not exceeding 17 .5 ksi. There are also a number of limitations on geometric parameters, consistent with the available test data used for the preparation of the WRC Bulletin. The basic philosophy behind nozzle reinforcement rules is to assure that the burst pressure of the vessel with the opening is at least equal to that of a vessel without any opening. Test results summarized in the WRC Bulletin 335 indicated that full area replacement was not needed to achieve this goal. Based on test results, rules were developed so that with reinforcement considerably less than full area replacement the burst pressure of the cylindrical vessels could be restored. The analysis of test results was based on the assumption that the vessel is subjected to design pressures allowed by present VIII-1 criteria. For a relatively small increase in allowable stresses, the burst pressure of the opening will probably still maintain a safe ratio to the burst pressure of the penetrated shell. However, since the stress distribution around the reinforced opening is not uniform, the ratio of burst pressures for penetrated and unpenetrated shell does not remain constant. If design margins are reduced beyond that recommended in phase 1 of this study (margin on tensile strength of 3.5), the provisions of this Code Case will have to be re-examined. One of the conditions of the Code Case is that "cyclic loading is not a controlling design requirement." This was in recognition of the fact that there is no attempt to maintain the same fatigue life at a reinforced opening as that of the unpenetrated shell. In fact, due to high local stresses and stress raisers, a reinforced opening almost always is inferior to the unpenetrated shell from a fatigue standpoint. Allowing higher design pressures will accentuate this problem. Even though VIII-1 does not explicitly address the fatigue problem, the threshold of the number of cycles for vessels for which no fatigue analysis is performed will be lowered. Any decrease in design margins will have to take this into account. E. C. Rodabaugh is in the process of preparing a report for the PVRC which will address the difference betweenASME codes and European codes. This study will include a detailed evaluation of Code Case 2168. The comparisons in this report are in the form of all reinforcement being placed in the nozzle neck. Comparisons of this report which include VIII-1 rules [including 1-7(a) requirements], VIII-2 (including special requirements of AD-560), Code Case 2168, and BS5500, indicate a very wide range of results for various geometric parameters. It indicates that reinforcement requirements of Code Case 2168 are considerably less than those of VIII-1 and VIII-2, but closer to those of BS5500. These results reiterate the fact the VIII-1 and VIII-2 reinforcement rules are more conservative than those of BS5500, but Code Case rules are generally less conservative than BS5500. Various codes put different limits on the size of an opening above which reinforcement is required. The rules ofVIII-1 have a nozzle diameter of 2%" to 3%", based on shell thickness, but regardless of vessel diameter. Tl:ft!'rules ofVIII-2, put this limit at a ratio of dl{DT exceeding 0.14. The limit on this ratio in BS5500 is 0.10. Code Case 2168 does not have a limit on the size of openings exempt from reinforcement, but since this is an alternative to basic VIII-1 reinforcement rules at least the same limits are applicable. The rules of the Code Case may result in no reinforcement requirement for penetrations larger than those exempt by the rules ofUG-36, depending on vessel geometry. In terms of a multiple of dl{DT, E. C. Rodabaugh's forthcoming report indicates that nozzles with ratio up to about 0.3 would be exempt. Evaluation of Design Margins, Section VIII 27 There is no question that an unreinforced opening, no matter how small, will result in stress concentration. Even though the effect of such small unreinforced openings may be insignificant on burst pressure, they definitely degrade fatigue capability of a vessel. With higher allowable stresses in vessel shells, the effect of unreinforced openings on vessel fatigue life will have to be addressed. 4. The Effe.ct of Lower Design Margins on Brittle Fracture in Part UHT Materials The Report by Dr. S. Yukawa on Brittle Fracture Margins in ASME Code Section VIII, Division 127 evaluated the margin for 38 ksi and 50 ksi yield strength materials based on the ratio of the available fracture toughness Krd to the applied Kr, where Krd was calculated from the Barsom-Rolfe Km-CVN correlation for dynamic fracture toughness, Krd = ~5E(CVN). The applied Kr was evaluated using the Section VIII, Division I allowable stresses and the maximum allowable flaw sizes in Section VIII, Division 1 for radiographic examination. Dr. Yukawa's evaluation of the Section VIII, Division 1 Code margins for resistance to brittle fracture in carbon and low alloy steels was based on: The margin on brittle fracture for Division 1, Part UHT high strength, quenched and tempered steel SA 517 Gr. F was evaluated in this report by determining the ratio of the available fracture toughness to the applied K for the Code permissible flaw sizes for radiographic examination (Section VIII, Division 1, UW-51). High strength steels inASME Section VIII (Su 2:: 95 ksi) must meet at least 15 mils lateral expansion. Recent revisions to both Division 1 and Division 2 increase the MLE requirements for nominal plate thickness exceeding 11;4". 10 The following Charpy energy values were obtained from data provided by Dr. W. D. Doty34 for SA-517 Gr. F steel with typical yield strength of about 110-120 ksi and from Fig. A2-2 and A2-3 in WRC Bulletin 175 14 (Table 1): Mils Lateral Expansion, MLE Estimated Charpy Energy (CVN), ft-lbs 15 20 25 35 23 30 40 63 1. Critical vs. permissible flaw size for a given stress, and 2. Material fracture toughness vs. applied stress intensity factor, Kr, for a given flaw size. The report by Dr. Yukawa gives the following results for Section VIII, Division 1 vessels: • Calculated critical flaw length vs. maximum Code permissible flaw length of %" for RT examination, for various flaw depth/thickness (alt) ratios and applied stress/yield strength ratios (SR). The size margin for SR 1.0 and alt = 0.5 is about 2% in 1-3 in. thick material with 38 ksi minimum yield strength (and higher for lower alt ratios and lower stress ratios). • Available fracture toughness (based on minimum required toughness in the Code) vs. the applied stress intensity factor K for a flaw with a length of%" (maximum permitted in UW-51 for RT examination). This evaluation indicated a minimum margin of about 1.5 for 1" and 3" thicknesses for a material with a minimum yield strength of 38 ksi and a stress ratio CSR) of 1.0 (and higher values for lower stress ratios). • The report also states that the relief from notch toughness testing provided in Division 1 (UCS66.1) by the reduced stress criterion does not compromise these margins. The maximum allowable defects for radiographic examination are given in UW-51 of ASME Section VIII, Division 1 and in Article I-5 of Division 2. Any elongated inclusion, such as slag, is limited to: • 114" fort up to%" (where t =thickness of the weld) 28 Table 2 • Ya t for t from %" to 2114" • %" for t over 2114" Static plane strain fracture toughness Kic and the yield strength of weld Sy, (assumed to be equal to th specified minimum yield strength of the base metal were used for these fracture mechanics calculations The Krc values were obtained from the lower boun Krc - CVN correlation, Kic = 9.35(CVN) 0 ·6 s. 35 ,3 6 Th results obtained with ASME Section XI, Articl A-3000 fracture mechanics procedures for %", 1114 and 2:Y2" thick SA 517 Gr. F are given in Table 2 These results are based on Section VIII, Division and 2 allowable defect lengths for radiographic exami nation and a surface flaw across the welded join with a depth to length ratio (all) of% (which resulte in the highest value of Kr). Based on the above calculations, the margin o brittle fracture is about 1.5 for the %" thick materi and about 1.2 for the 1114" material in the as-welde condition (using the UW-51 maximum permissibl flaw sizes for radiographic examination and assum ing S = Sy for the weld, acting uniformly across th thiclmess). Although there is a significant amount o successful experience with the present Code tough ness requirements, a further review of the Cod toughness requirements for these steels is sug gested, particularly for non-stress relieved welde joints over a certain thickness (e.g.%"), regardless o the design stress. In the stress relieved condition with the desi stress S = Su/4 and residual stress equal to 30% o the yield strength of the weld, the margins on brittl WRC Bulletin 435 S S S S S S 2W' 22.5 (35 ft-lbs) 87.8 0.75" S S S S =Sy= 100* Su/4 28.8 Su/3.5 = 32.9 = Sy 100 ksi* = Su/4 + 0.3Sy = 58.8** (PWHT) = Su/3.5 + 0.3Sy = 62.9** (PWHT) Su/4 28.8 ksi = Su/3.5 = 32.9 = Su/4 + 0.3Sy = 58.8** Su/3.5 + 0.3Sy = 62.9** = 44.2 15.6 17.9 57.1 32.4 1.5 4.3 3.8 1.2 2.1 34.7 1.9 20.9 24 43.3 46.5 4.2 3.7 2.0 1.9 Notes: *This represents a non-stress relieved butt weld with the design membrane stress plus welding residual stress equal to the yield strength of the weld (Sy). **This represents a stress relieved (PWHT) butt weld with the assumed remaining residual stress after PWHT equal to 30% of the room temperature yield strength of the weld in the direction of the weld (26, 36). (ASME Code requires PWHT ofSA-517 welds at 1000°F-1100°F for thickness t > 0.58'', or fort> 1Y4" when preheated 200°F and post-heated 400°F over 0.58"). Sy = specified minimum yield strength. Su = specs minimum fenride strength. fracture are about 2.1 for the 1114" thick welded joints and 2.0 for the 2%" welded joints. Decreasing the design margin from 4 to 3.5 (by 14.3%) slightly decreases the margin on brittle fracture. In the stress relieved, the margins on brittle fracture are about 1.9 for the 1114" and the 2%" thick welds welds. Gross structural discontinuities and stress concentrations at the edge of the weld, however, can significantly increase local secondary and peak stresses in the direction transverse to the weld and thereby increase the risk of brittle fracture of flaws which may exist in the weld or heat affected zones in a plane parallel to the weld. Precautions, therefore, need to be taken in high strength steels to avoid structural discontinuities (such as misalignment and peaking) and stress concentrations at 101al structural discontinuities (such as notches, sharp corners at edges of the weld, etc.). More sensitive NDE techniques, such as wet magnetic particle surface examination of welds and/or UT, provide additional means of detecting defects in high stress areas. The risk of brittle fracture often is the highest during overload testing of the vessel, because the vessel is being subjected to high stresses for the first time. It is, therefore, important to assure adequate margin against brittle fracture for both the overload test and for the design/operating conditions of the vessel. Section VIII, Division 1 presently recommends that the minimum metal temperature during the hydrostatic test be at least 30°F above the minimum metal design temperature. It is a Code requirement for the pneumatic test. Consideration should be given to making the 30°F differential mandatory also for the hydrostatic test when a lower design margin is adopted. D. Study of Vessel Test Results 1. Purpose This section of the report presents a summary of available pressure vessel test data. Based on the test results, some conclusions are made regarding the safety of fabricated vessels designed to various rules. 2. CBI Tests of Pressure Vessels with Reduced Thickness Girth Seams In 1987, CBI performed destructive testing of 10 pressure vessel models to justify the use of reduced thickness girths, without reducing the allowable pressure. This concept takes advantage of the fact that, in cylindrical shells, the axial stress is one half of the circumferential stress. The results of this project were used to introduce a Code Case in VIII-2 to allow reduction in girth weld thickness, without a reduction in allowable pressure. This Code Case has now been incorporated in VIII-2, as paragraph AD200(b). The test models included a reference shell without any reduction in thickness and other models with varying degrees of reduction in thiclmess and various beveling to the reduced thickness. The material for the cylinders was 20 inch diameter schedule 80 seamless pipe (SA-333, Gr. 6). The calculated allowable design pressure, based on VIII-2 rules, was 1320""si. The actual burst pressures ranged from 4276 to 4459 psi. The vessels, having been egg shaped at failure, split along a longitudinal seam. This of course proved that the reduction in a girth thickness has no appreciable effect on failure pressure of a cylindrical shell. But these tests also indicated that margins to failure for vessels fabricated to VIII-2 rules is in the order of 3.24 to 3.38 for a ductile failure of cylindrical shells of carbon steel vessels with a diameter/thickness (d/t) ratio of about 17.4. If VIII-1 allowables were used these margins would have been 4.32 to 4.50. This of course is for a Evaluation of Design Margins, Section VIII 29 clean shell, without any penetrations or attachments, constructed of a notch tough material (the measured expansion in diameters ranged from 25 .1 % to 34.5%). The details of these tests are included in a CBI report. This brief explanation points to the safety of cylindrical vessels fabricated to ASME Code rules. 3. Ellipsoidal Head Model Tests In 1984, CBI performed pressure tests to failure of two models of nuclear containment vessel heads. 21 The two torispherical heads, fabricated from SA-516 Gr. 70 material, had cylindrical diameters of 192 in., knuckle radii of 32.64 in., spherical dome radii of 172.8 in., and nominal thicknesses of 3/iG and Y-4 in. For the o/16 in. thick head, initial buckling was noted at 58 psi. Further pressurization rounded out the head until rupture occurred in the knuckle at 229 psi. For the Y-4 in. thick model, initial buckling was noted at 106 psi and rupture occurred at the top of the head at 332 psi. The allowable design pressures calculated by the rules of Subsection NE of Section III ASME Code were 33.2 psi and 45.5 for models 1 and 2. Using VIII-1 rules, the allowable design pressures would have been 36.5 and 50.1 for the two models. This would result in calculated margins to failure of 6.27 and 6.62. These results indicate the large margins available on rupture of torispherical heads. For these thin heads, of course, the failure mode is considered to be knuckle buckling, with resulting margins which are much lower. It should also be pointed out that these observed margins to rupture may not hold over the range of geometric parameters. These tests give an indication of safety of thin torispherical heads. 21 allowables. This indicates a considerably higher margin for this larger and thicker vessel. 5. Praxair Torispherical Head Test In 1994, Praxair performed pressure testing of two torispherical heads for the PVRC. The purpose of these tests were to evaluate the safety margins in the ASME Code rules for design of formed heads and to provide a basis for developing new and updated rules. Both heads were attached to cylinders with 60 in. diameters and were made of SA-516 Gr. 70 material. The crown radius-to-diameter ratios were 1.0 and the knuckle ratios were 6%. The nominal head thicknesses were 0.25 in. and 0.32 in. The allowable design pressure, using the rules of VIII-1 were 85 psi and 110 psi. For Model 1, the pressure was increased to 700 psi at which time a penetration on the attached cylinder failed. Model 2 failed at 1080 psi by ductile fracture along a longitudinal seam of the attached cylinder. These failure pressures indicate minimum margins for the heads of 8.2 and 9.8, respectively. This confirms the widely held belief that the presentASME Code rules for design of formed heads is overly conservative, at least over a certain range of parameters. 6. MPC Service Exposed Test Vessels With Local Thin Areas The objective of this test program was to evaluate during a hydrostatic test the performance of two service exposed, retired pressure vessels with deliberately prepared local thin areas (LTA's) of varying geometry and locations. The vessels were originally designed to ASME, Section VIII, Division 1. 3l The vessel dimensions, material properties, wall thickness and design conditions are listed below (Table 3): 4. Containment Vessel Model Tests A series of tests were performed in the mid-80's by Sandia National Labs, to estimate the margin to failure of nuclear containment vessels. Four models had a diameter of 43 in. and thickness of .045", with varying details; some including major penetrations: These models were fabricated to the rules of Section III, Subsection NE, and had calculated allowable design pressures of 40 psi. The failure (rupture) pressures were 140 psi, 110 psi, 120 psi, and 120 psi. The calculated margins to failure, using Subsection NE rules were 3.5, 2.75, 3.0, and 3.0 respectively. If the design rules of VIII-1 had been employed the margins to failure would have been 3.85, 3.02, 3.3 and 3.3. This indicated a lower margin than expected. The fact that these models were fabricated of thin gage material and included major discontinuities probably contributed to these lower than expected margins. This test program also included a larger model with diameter of 168 in. The calculated design pressure for this vessel, using Subsection NE rules, was also 40 psi and the burst pressure was 195 psi. This margin of 4.88 would have been 5.37 for VIII-1 30 Table3 Inside diameter Length Wall thickness Design pressure Test pressure Design temperature Material Shell nozzles/Manways vessel 1 Vessel2 48" 144" 7/16" 125 psi 297 psi 600°F SA-285, Gr. C 4" & 18" diameter with reinforcing 78" 199" l" shell, 1.25" heads 300 psi 523 psi 650°F SA-515, Gr. 65 4" & 18" diameter with reinforcing Vessel 1. The vessel had significant internal corro-, sion, which was removed by sandblasting. The aver~ age wall thickness of the vessel wall ranged froni 0.30 to 0.41 inches, and a nominal wall thickness was assumed for this vessel for the purpose of this project. There was also a significant amount o blistering, as a consequence of wet H 2S service About 30 local thin areas with tapered edges wer ground into the vessel shell and heads. The size an depth of the individual LTA's were selected such tha the remaining strength factor (RSF) associated wit WRC Bulletin 435 each LTA would be about 0.8. The intent was to produce several pads with a net thickness of about o.1" at the bottom of the LTA. The LTA's were positioned at various locations on the vessel, including edges of the nozzle reinforcing pads, but away from areas which were heavily corroded or blistered. At test pressure of 515 psi (average hoop stress of about 40,000 psi in 0.31" thick shell) a leak occurred in one LTA, under a strain gage, which had a strain gage reading that reached 5.9%. Several other LTA's had strains ranging from 2.1 % to 5.9%. The vessel wall did reach nominal yield stress. No deformation was reported at nozzles or manholes. The test temperature was 55°F. Vessel 2. The as received vessel was in a better condition than Vessel 1. There was little or no corrosion and blistering. To generate about the same remaining strength (RSF) as for Vessel 1 at the LTA's, the wall thickness at some LTA's was reduced to about 0.25". The locations of the LTA's were similar to Vessel 1, except that some additional LTA's were located at nozzle-to-shell welds and along a longitudinal weld seam. Also, some long, narrow grooves were ground into the shell in the axial direction. After 1050 psi test pressure some of the strain gages in the LTA's began to follow a previously established strain hardening curve. Nearly all strain gages in the hoop direction had significant plastic deformation. The test was terminated at 1315 psi (about 51,300 psi nominal hoop stress in the 1" thick vessel shell) due to excessive leakage at a nozzle gasket. The nominal strain gage reading at a test pressure of 1300 psi was 0.69%, indicating general yielding in the vessel wall. The maximum local strain reading at an LTA was 4.69%. No deformation was reported at nozzles or manholes. The test temperature was 84 °F for Vessel 2. Vessel 1 had a margin of 2.91 on rupture and Vessel 2 test was terminated at a stress equal to 3.16 times the design stress, based on the nominal hobp stress in the vessel shell at the end of the test. This does not account for the RSF of about 0.8 in the LTA's. No information was provided on the actual tensile properties of the materials in these test vessels. E. Evaluation of Design Margins for ASME Section VIII, Division 2 1. Review of Failure Modes Considered in Section VIII, Division 2 In 1987-1989 the ASME Subgroup on Design Analysis (SGDA) performed a study to determine the failure modes considered in the various sections of the ASME Code and the intended safety margins for the given failure mode in each Code section. 25 Eight failure modes were identified and discussed in the February, 1989 report "Summary of the Design Analysis Factors Inherent in the Established Failure Modes of the ASME boiler and Pressure Vessel code," prepared by SGDA. The following is a brief summary on each failure mode. A description of design margins explicitly introduced into the Code design rules is included in ASME Section II, Part D, Appendix2. a. Ductile Rupture ASME Section VIII, Division 2 uses the Tresca, or maximum shear stress criteria to establish the allowable stress intensities. The shear stress is equal to one half of the difference between any two principal stresses at any given point in a structure. Yielding will occur when the difference between any two principal stresses, defined as a stress intensity, S, equals the yield strength. The stress intensity in Section VIII, Division 2 is the maximum allowable value of the stress difference. Appendix 2 in Section II, Part D gives the following criteria for establishing design stress intensity values: Table AD-150. l gives the following k factors for establishing the maximum stress intensity limits, k Sm, for various load combinations (Table 4): Table4 Design (Design pressure, dead load of the vessel, vessel contents, external attachment loads) Design + wind Design + earthquake Design + wave action Hydrostatic test Pneumatic test 1.0 1.2 1.2 1.2 1.25 1.15 7. Conclusions The available test results indicate that for cylindrical vessels, the safety margins against rupture are approximately the same as design factors on ultimate tensile strength. The test vessels designed to Section VIII, Division 1 failed at design margins ranging from about 3.0 to 4.5, with an average of about 3.7 on the tensile strength of the material. For formed heads, the safety margins to rupture are c?nsiderably higher. This of course is a generalization. The data are very limited and does not cover the entire range of parameters. The true safety margins are also a function of fabrication details, quality of welds and many other factors, which cannot be covered by a finite number of test models. The basic stress intensity limits are listed in Table 5. Tables General primary membrane <Pml Local primary membrane (Pd Primary local membrane + primary bending (PL + Pb) + secondary k Sm :s k (lower of 2/3 Sy or ST/3) 1.5 k Sm 1.5 k Sm 3 Sm + ST = specified minimum tensile strength at room temperature. Sy specified minimum yield strength at room temperature. Evaluation of Design Margins, Section VIII 31 Table 6 lists the ASME Section II, Part D, Appendix 2 criteria for establishing design stress intensity values for ferrous and non-ferrous materials (other than bolting) at temperatures below creep range. The intent of these design margins is to provide adequate margins of safety at temperatures below the creep range against the following: • Ductile rupture • Local failure • Uncontrolled bending deformation A nominal design margin of 3 is provided against ductile rupture. The design margin can be as low as 2.5 under some operational loading conditions and 2.4 under the hydrostatic test condition. The local design margin on ultimate tensile strength can be as low as 1.6 for the hydrostatic test condition. (True margins of safety generally are greater, since local stresses in the vessel can redistribute.) b. Brittle Fracture The toughness requirements for ferrous materials is covered in AM-210 of Section VIII, Division 2. Toughness requirements for quenched and tempered steels are included in AM-310. Recently approved changes to the Division 2 toughness rules have made them essentially the same as Division 1 toughness requirements. In both Divisions the materials are grouped into four groups based on the notch toughness characteristics of the materials. Fig. AM-218.1 shows impact test exemption curves A through D for materials in the four groups of carbon and low alloy steels. Curve A includes the least notch tough steels, including structural shapes and bars. Curve D includes heat treated fine grain steels with improved toughness. Impact testing is required at minimum design metal temperature/thickness combinations below the exemption curves and for thicknesses above 4 inches. The required minimum average impact energy values are given in Fig. AM-211 of the revised toughness rules and depend on thickness and the specified minimum yield strength of the material. The orientation of the impact test specimen is not specified. Fig. AM-218.2 of the revised toughness rules provides for a reduction of the impact test temperature based on the ratio of the required thickness to the nominal thickness of the component. All Table ACS-1 carbon and low alloy steels with specified minimum ultimate tensile strength >95 ksi and all Table AQT-1 high strength quenched and tempered steels shall meet 15 mils minimum lateral expansion for thicknesses up to 1114" and 25 mils for thicknesses above 3" (Fig. AM-211.2). Other Section VIII, Division 2 requirements relevant to avoidance of brittle fracture are: • The Code requires impact testing of welds and RAZ when base metals are required to be impact tested (Article T-2). • Essentially all welded joints required to be non-destructively examined (TableAF-241.1). he Section VIII, Divisions 1 and 2 toughness rules are based on fracture mechanics considerations and extensive good experience with ASME Code vessels. It is not possible, however, to establish a single margin of safety against brittle fracture in the Code rules because brittle fracture depends on three factors, the applied stress, the flaw dimensions, and the available toughness of the material. c. Fatigue AD-160 in Section VIII, Division 2 gives rules for determining exemption from fatigue analysis. The rules for exemption from fatigue analysis are based on experience with low cycle operation. The procedures for design for cyclic loading are included in Appendix 5 of the Code. The fatigue curves in Appendix 5 are limited to a temperature of 700°F for ferritic and 800°F for austenitic steels. The fatigue design curves in Article 5-1 were derived from fatigue tests on smooth specimens tested at room temperature. Peak stresses are calculated by assuming or evaluating stress concentration factors at discontinuities and comparing these stresses to the smooth speci- men design S-N curve. TheASME fatigue design curves include a factor of 2 on stress in the high cycle regime and 20 on cycles in the low cycle regime (approximately <10000 cycles). The factors account for scatter of the test data, size effects, surface finish, and industrial atmosphere. d. Incremental Collapse The Code imposes the following limitations on the Primary + Secondary Stress Intensity to provide a margin against progressive deformation during cyclic loading: The 3 Sm limit is equal to or less than 2 Sy. The 2 S limit results in elastic cycling after initial yieldin~ (shakedown). e. Elastic Instability The basis for buckling rules is outlined in Appendix 3 of Section II, Part D. The true safety factor built into these rules varies depending on loading and the geometry being considered. The value of applied safety factor ranges from 2.0 to 4.0. This nonuniform safety factor is one of the shortcomings of these rules. In addition to safety factors, knockdown factors from test results have been applied to the theoretical buckling stresses. These knockdown factors, being based on the lower bound of test data provide additional safety margins. The maximu~ buckling stresses are limited to the allowable tensile stresses. In addition, for cylindrical shells, these stresses are limited to Ys of minimum yield strength. f. Plastic Instability The Code imposes the following limitations on stress intensities to avoid plastic instability: 1. General primary membrane stress Pm :5 % (Appendix 4) 2. Local primary membrane stress, PL :::; Sy 3. Local primary membrane plus primary bending stress, PL + Pb :5 Sy. Sy = specified minimum yield strength at tempera- ture. Welded pipe or tube, ferrous or non-ferrous The purpose of these design margins is to provide a margin against instability where tensile straining by a primary load can produce large deformations. 1.1 Wrought or cast, ferrous or non-ferrous 3ST~ 0.85 (l.lx0.85) ----ST~ 3 0.85 LlSy 0.85 LlSyRy g. Excessive Deformation and Leakage The Code does not have explicit rules for determination of deflections and rotations. For design of 0.92 (l.lx0.92) 0.92 Ferrous materials, structural quality, pressure flanges, Division 2 requires that the stress allow----ST~ LlSy - -ST retention 3 3 a~les of Division 1 be used, in an attempt to miniFerrous materials, structural quality, non-pres'l:i ST 1.1 % Sy % SyRy or 0.9 m1ze the probability ofleakage. Division 1 does have sure retention 3 ST~ SyRy - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - " , I l l ~on-mandatory rules in Appendix 5 for further limitmg the rotation of closure flanges, but Division 2 ~=ratio of trend curve value of tensile strength to room temperature tensile strength. does not have such rules. Ry= ratio of trend curve value of yield strength to room temperature yield strength. 32 WRC Bulletin 435 or (0.9 x 0.85) SyRy 0.92 LlSyRy 2. Assessment of Significant Factors on Design Margins a. Design Details Design details have a significant effect on the true margin of safety for pressure vessels. However, it is difficult to quantify the effects of various details on d~ff~rent fa~lure modes. For plastic collapse, an analysis is reqmred. For fatigue, a stress concentration (fatigue stress reduction factor) may be calculated or assumed. For brittle fracture, the effect of local stresses and the notch effects of various details need to be taken into account. Details that are allowed are normally based on judgment and experience. For some details, test results and/or analysis results are available. All these resources need to be taken advantage of in establishing acceptable details. b. Ductile Rupture !he limits on local membrane stress intensity and membrane plus primary bending stress mtens1ty (1.5 Sm) in ASME Section VIII, Division 2 have been set at a level which conservatively assures the prevention collapse as determined by the principles of limit analysis (Paragraph 4-136(b)). Since the primary stress intensity limits are based on factors of yield strength and the ultimate tensile strength of the material, it is generally understood that this provides assurance against excessive distortion or bursting mode of failure. B. F. Langer 17 lists the following modes of failures related to bursting at temperatures below the creep range: ~r1ma~y 2. Bursting due to general yielding and rupture of the vessel shell 3. Ductile tearing at discontinuity such as a nozzle, or head-to-shell intersection 4. Brittle fracture at a defect Langer 12 recommended the use of the modified Svensson equation for calculating the maximum burst pressure based on plastic instability since that correlated well with experiments. The modified Svensson equations are: PB = O'u Fey! ln W, for cylindrical shells PB = O'u Fsph ln W, for spherical shells O'u ultimate tensile strength of the material Fey! [0.250/(Eu + 0.227)] (elEu)eu, correction factor for a cylinder and represents the effect o:rt'train hardening on burst pressure, Fsph = [0.250/(Eu + 0.227)] (l.155e1Eu)eu, correction factor for a sphere W = ratio of the vessel outside diameter (d 0 ) to the inside diameter (dJ For relatively thin-walled vessels, the In W can be approximated by ln W = 2t/(di + t) E = true strain e = logarithmic base Eu = ln(l + E~), true strain at maximum load in the tensile test (true maximum elongation) and is used in the modified Svensson equa- Evaluation of Design Margins, Section VIII 33 0" 0 E'u tion as an approximation of the strain hardening exponent, n, for materials whose true stress-strain curve is represented by O" = 0"0 En = strength coefficient (true stress at E = 1.0) = engineering strain at maximum load (uniform elongation) The relationship between Fey! and Fsph and Eu are presented graphically in Fig. 1. Fey! and Fsph are numerically greater for lower values of Eu· The burst pressure is proportional to F cyl and F srh and, therefore, higher for vessels made from high strength steels because of increased tensile strength and the lower strain hardening exponent. 2.0 1.9 ~ ' 1.8 1.7 1.6 ' "," ~Foph ' 1.5 1.4 " ""- cc ~ 1.3 u < LL. LL. "' "~ ~ 1.2 I.I 1.0 ' ....... '-.... "'" 0.9 0.8 ........ ~,Fcyl ........... ............_ 0.7 ............. !'-....... 0.6 0.5 0 .2 .3 .4 .5 .6 .7 .8 ............. .9 t.0 E:u Fig. 1-Values of cylinder and sphere factors for modified Svensson equation Langer used the modified Svensson equations to predict the actual margins against bursting mode of failure in spheres and cylinders of four different materials and d/t 10, designed to Section VIII, Division 2 allowable stresses. 17 The margins against bursting mode of failure (SF), calculated by Langer, are listed in Table 7. The results in this table indicate that the predicted margins on bursting pressure (S.F.) are somewhat higher for cylinders than for spheres with the same d/t. PVRC Subcommittee on Effective Utilization of Yield Strength sponsored several projects in 1960's 34 SA-240, 830400 SA-516, Gr. 55 SA-533, Gr. B SA-517, Gr. F 75 55 80 115 30 30 50 100 20 18.3 26.7 38.3 0.55 0.31 0.17 0.11 2.73 2.49 2.69 2.79 2.91 2.75 3.03 3.17 and early 1970's to provide a better understanding of the behavior of pressure vessels in the bursting mode of failure and to evaluate the accuracy of the modified Svensson equation. 19·20 The project at University of Kansas included burst tests on test vessels of three types of materials with different strain hardening exponents. The materials were SA-240 Type 304 stainless steel, SA-516 Grade 70 carbon steel and SA-517 Grade F quenched and tempered steel. These steels have strain hardening exponents of 0.585, 0.189, and 0.085, respectively. The burst tests included two types of strain concentrations: • Vessels with "moderate" strain concentrations common to pressure vessels, such as nozzles, welds, and end closures. • Severe strain concentrations, such as sharp notches in the axial direction of the test vessels: The notches were machined to depths about 15; 25, and 35% of the vessel wall thickness and had a root radius at or below 0.001" (0.025 mm). The PVRC sponsored work led to the following conclusions: 1. The yield strength controls the design of Divi~ sion 2 vessels only if the yield strength/tensile strength ratio is less than 0.5. Design ofSectio VIII vessels is essentially based on the burstin mode of failure, instead of yielding, in th temperature range in which bursting, rathe than the other failure modes (fatigue, brittl fracture, etc.) is the controlling criterion. 2. The modified Svensson formulas can be ex pected in most cases to predict the bursti pressure of a vessel to within about ± 10<7£ providing the material is sufficiently ductile a the time of failure to fail in a bursting mode o failure and not by brittle fracture. 3. The modified Svensson equation gives reason ably accurate prediction of burst pressures i thin wall pressure vessels with moderate strai concentrations. (Moderate strain concentra. tions did not significantly reduce the calculate burst pressure in the University of Kansas tes vessels with nozzles, welds and end closures). 4. For vessels with sharp notches, the test resul showed that a direct correlation between redu tion in wall thickness (due to a sharp not and reduction in burst pressure exists at leas until the notch depth exceeds 25% of the wa thickness for the steels tested. WRC Bulletin 435 The actual burst pressures for the University of Kansas test vessels with full wall thiclmess (without sharp notches) and the theoretical margins against the bursting mode of failure for these vessels using the modified Svensson formula for cylinder are listed in Table 8. The above predicted bursting pressures are based on the specified minimum tensile strength values, whereas the actual bursting pressures are based on the actual measured tensile values, therefore, the actual factors of safety are somewhat higher than predicted. ~aln~ns and Updike discuss the safety margins of cylmdrical. and spherical shells against bursting mode of failure without and with edge restraints.28 Ba~ed ?n their anal~sis of the plastic instability of cylmdncal shells with end closures they conclude that restrained cylindrical and spherical shells burst at higher pressures than unrestrained shells.29 Fig. 2 shows the margins against bursting mode of failure vs. par8:meter UR (L = length of cylindrical shell, R.= radm~) for a cylindrical shell that is designed with a pnmary membrane stress Sm = Su/3. The asymptote for a large UR approaches a safety factor s.s;1 - , - - , - - , - - . - - , . - - - - . - - - , - - - - - ' : 5 I.JR Fig. 2-Safety factors for restrained cylindrical shells of 3.2 on bursting pressure for the SA-516. Gr. 70 material. Some European pressure vessel codes use lower margins on the tensile strength than ASME. The British BS 5500 uses a factor of 1.5 on yield strength and 2.35 on tensile strength for carbon and low alloy steel vessels. The CEN draft Standard for Unfired Pressure Vessels proposes to use the lower of the following design stresses for other than austenitic steel for pressure parts other than bolts: • The minimum specified yield strength or 0.2% proof strength (Sy) divided by a safety factor of 1.5. • T~~ minimum specified tensile strength (Su) dlVlded by a safety factor of 2.4. (The July 1995 draft of the CEN/TC draft Standard for Unfired Pressure Vessels included an alternative design basis with a design stress based on the lower of Su/1.875 and% Sy. However, this design basis is not included in the May 1998 draft). Tl~e predicted margins against the bursting mode of failure for the cylindrical vessels tested at University of Kansas were re-calculated using the modified Svensson equation and allowable design stresses based on the lower of Su/2.4 or % Sy and are listed in Table 9. Based on the same allowable stresses (the lower of Su/2.4 or % Sy) spherical vessels with the same inside diameters and thicknesses as the University of Kansas cylindrical test vessels have somewhat lower margins than cylinders on the predicted burst pressure. The predicted margins are given in Table 10. ~though this results in lower predicted margins agamst the burst pressure than vessels designed to ASME Section VIII, Division 2, there have been no indication o&bursting mode of failures, therefore, it can be concluded that the European codes have shown adequate safety in practice against this failure mode. Evaluation of Design Margins, Section VIII 35 lent Kie values for this study are listed in the following table (Table 11): Table 11 ASMECode c. Brittle Fracture In fracture mechanics calculations, the value of the stress intensity factor, Kr, is based on the applied stress and dimensions of the flaw. The calculated Kr must be less than the critical fracture toughness parameter Krc (or Krd, depending on the loading rate) to avoid brittle fracture. Also, consideration must be given to both primary and secondary stresses to evaluate maximum stress intensity, Kr, in the structure. This part of the report attempts to evaluate the following for vessels of several ASME Section II materials (SA-516 Gr. 70, SA537 Cl. 1 and Cl. 2, and SA517 Gr. F and Gr. E): 1. The change (reduction) in critical sizes (depths) of surface flaws with lower design margins (higher design stresses) but with the same toughness requirements as presently in ASME Section VIII, Division 1 and Division 2. 2. The toughness needed to retain the same margins on brittle fracture (the same critical flaw size) for vessels designed to higher stresses as for vessels designed to present Code allowable design stresses. The following cases were investigated: • Increasing the design stress from Su/4 to Su/3.5 • Increasing the design stress from Su/3 to the lower of Su/2.4 or% Sy SA-537,Cl. l SA-537,Cl.2 SA-517, Gr. E &F SA-517, Gr. E 51.5 51.5 57.8 57.8 51.5 59.8 66.5 66.5 61.7 69.2 79.7 79.7 67.4 78.0 95.5 95.5 Ba~e Metal. The following cases list the design margms selected for this study: Case 1: S = Su/4 Case 2: S = Su/3.5 Case 3: S = Su/3 Table 12 lists the materials and stress values (ksi) for this study. ~he critical depths of surface flaws with an aspect rat10 of all= 0.15 (l/a = 6%) were obtained for Cases 1-~ with the PD 6493:1991, Level 2 FAD procedure. Tlus aspect ratio was selected from the computer printouts for convenience, mainly because it is close to the aspect ratio for the flaw dimensions specified in ASME Code Section III, Part NB(~ t x 1.5 t smface flaw). The calculated flaw depths are listed in Table 13. Table 14 indicates when leak-before-break (as defined by through thickness flaw with depth a = t and length 1 = 2t) was achieved for Cases 1-5 with th~ ASME Section VIII toughness requirements. !ills shows that leak-before break is possible practically 111 ~l the cases examined in the above table with the Sect10n VIII, .Division 1 design stresses (Su/4) and in most cases with a design stress of Su/3.5. None of the materials investigated (except possibly in lesser thicknesses) have leak-before-break capability when the maximum primary plus secondary stresses are equal to Sy. Tables 15 and 16 show the toughness needed for Table 12 Material SA-516, Gr. 70 SA-537, Cl. l SA-537, CL 2 55 MILS LATERAL EXP. 15 ft-lbs 15 ft-lbs 18ft-lbs 18 ft-lbs 15 ft-lbs 19 ft-lbs 22.5 ft-lbs 22.5 ft-lbs 20 ft-lbs 24 ft-lbs 30ft-lbs 30ft-lbs 15 MLE (=23 ft-lbs) 19.3 MLE (=29 ft-lbs) Thickness Case 1 Case2 Case3 Case4 S = Su/4 S = Su/3.5 S = Su/3 S:::; Su/2.4'1' Case5 S =Sy All ts21;;2'' 21;;2''< ts 4" ts 45 75 40 § ~ 35 >- Table 13 es Sy = specified minimum yield strength Su = specified minimum tensile strength ~ :x: ~ The assumed increase in design stresses represent the following decrease in the design margins(%): S = Su/4 to S = Su/3.5: 14.3%. S = Su/3 to lesser of S s Su/2.4 and% Sy: 8.6% for SA-516, Gr. 70 and 25% for the other steels included in these calculations. The fracture mechanics evaluation was done with the Fracture Graphic® Version 1.1 computer program which incorporates the British PD 6493:1991 failure assessment diagram (FAD) procedures, Levels 2 and 3. 26 Most of the calculations in this section are based on the Level 2 FAD procedure and the ASME Section VIII Division 1 and Division 2 notch toughness requirements. 36 The ASME Section VIII required Charpy energy values were obtained from Fig. UG-84.1 in Division 1 and the recently revised toughness rules in Division 2, Fig. AM-211. The Code toughness requirements for high strength steels, such as SA-517 Gr. F and Gr. E (Su ::::: 95 ksi), must meet at least 15 mils lateral expansion, or higher, depending on thickness. The minimum mils lateral expansion (MLE) values were obtained from the new Fig. AM-211.2 in Division 2.10 Since there are no known Kic - MLE correlations, CVN values for the SA-517 Gr. F and Gr. E steels were estimated from Fig. 3 (Fig. A2-1 in WRC Bulletin 175 14) and the MLE - CVN correlations 34 provided by Dr. W. D. Doty for SA-517 Gr. F. Kic values for the fracture mechanics calculation were obtained from the lower bound Kic - CVN correlation Kic = 9.35(CVN) 0·63 , to provide a high degree o conservatism. 35 •36 The materials, the thicknesses, the ASME Section VIII notch toughness requirements, and the equiva- 3" 4" 1%" 2" 3" 4" lY:i' 2" 3" 4" lW' 2" 3" 4" Case 4: S = lesser of Su/2.4 and % Sy 5: S = Sy, assuming that in some cases the prnnary plus secondary stresses could be as high as the specified yield strength of the material at local discontinuities. Cas~ 50 0 25 "F > ~< :x: <..,) 25 FOR PLATES. FORGINGS, WELD METAL AND WELD-HEAT-AFFECTED ZONES SA-537, Cl. l SA-537, CL 2 50 100 150 YIELD STRENGTH IKSll 1%" 2" 3" 4" 1%" 2" 3" 4" SA-517, Gr. F lY.," SA-517, Gr. E 2" 3" 1.018" 1.212" 1.680''. 1.940" 1.028" 1.451" 2.126" 2.496" 1.129" 1.472" 2.209" 2.601" 0.810" 1.188" 1.946" 0.883" 1.039" 1.429" 1.629" 0.899" 1.265" 1.883" 2.149" 0.857" 1.284" 1.928" 2.246" 0.70" 1.019" 1.467" 0.734" 0.849" 0.759" 1.064" 1.550" 1.775" 0.834" 1.083" 1.624" 1.866" 0.582" 0.840 1.397" 0.656" 0.752" 1.017" 1.122" 0.571" 0.795" 1.152" 1.283" 0.629" 0.812" 1.218" 1.364" 0.427" 0.605" 1.025" 0.287 0.312" 0.408" 0.427" 0.187" 0.256" 0.389" 0.406" 0.180" 0.229" 0.369" 0.382" 0.066" 0.090" 0.169" Fig. 3-Relationship at 25, 35, 40, 45 and 55 mils lateral expansi between energy and yield strength. (Fig. A2-1, WRC Bulletin 175). WRC Bulletin 435 37 Table 14 SA-516, Gr. 70 Yes Yes Yes 1=6.50" Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 1%" 2" 3" 4" SA-537, Cl.1 3" 4" SA-537, Cl. 2 SA-517, Gr. E & F SA-517, Gr. E 3" 4" lW' 2" 3" 4" 1 2.74"* 1 = 2.74"* 1 = 3.45"* No Yes 1 = 3.91"* 1 = 5.58"* 1 = 5.58"* Yes Yes Yes 1 = 6.06"* 1 = 1.86"* 1 2.49"* 1 = 4.48"* 1 4.47"* Yes Yes Yes 1 = 7.75"* 1= No No No 1= 1= 1 No 1 1 1= No No No No No 2.27"* No No No No No No No No No No No No No No No No 1.78"* 3.40"* 3.41"* 1.96"* 2.47"* 3.72"* Note: *This table also lists the calculated critical length 1of the through thickness flaw when t < 1 < 2t. Table 15 SA-516, Gr. 70 SA-537, Cl.1 SA-537, Cl. 2 SA-517, Gr. E & F SA-517, Gr. E lY2" 2" 3" 4" 1%" 2" 3" 4" lYz" 2" 3" 4" 1\1,i" 2" 3" 4" 15 ft-lbs 15 ft-lbs 18 ft-lbs 18 ft-lbs 15 ft-lbs 19 ft-lbs 22.5 ft-lbs 22.5 ft-lbs 20 ft-lbs 24ft-lbs 30 ft-lbs 30 ft-lbs 15 MLE (=23 ft-lbs) 19.3MLE 19.4 (=29 ft-lbs) 25 MLE (=40 ft-lbs) 25 MLE ( =40 ft-lbs) 63.5 (59.5) 61.7 (59.5) 68.6 (66.7) 67.9 (66.7) 61.4 (59.3) 70.9 (68.8) 78.8 (76.5) 77.8 (76.5) 73.4 (71.0) 82.0 (79.6) 94.6 (91.7) 93.3 (91.6) 78.5 (77.4) 90.5 (89.5) 23.3 (15.5) 19.8 (15.5) 18.7 (15.4) 17.5 (15.4) 19.2 (15.1) 18.6 (15.1) 18.5 (15.0) 17.0 (15.0) 19.0 (15.1) 18.5 (15.0) 18.7 (15.1) 17.1 (14.9) 16.5 (14.5) 16.0 (14.7) 21 (19) 20 (19) 23.5 (22.5) 23.5 (22.5) 20 (19) 25 (24) 29.5 (28) 29 (28) 26.5 (25) 31.5 (30) 39.5 (37.5) 38.5 (37.5) 29.5 (28.5) 37 (36) 111.3 (109.0) 110.7 (109.6) 16.5°(14.8) 15.9 (14.8) 51 (50) 50.5 (50) Table 16 Material SA-516, Gr. 70 SA-537, Cl. 1 SA-537, Cl. 2 Thickness, inches Sect. VIII, Div. 2 CVN Requirement Required Kzc,ksi,jin. %Increase inKzc Equiv. Charpy Energy, ft-lbs lYz'' 2" 3" 4" 1%" 2" 3" 4" lYz" 2" 3" 4" 15 ft-lbs 15 ft-lbs 18 ft-lbs 18 ft-lbs 15 ft-lbs 19 ft-lbs 22.5 ft-lbs 22.5 ft-lbs 20 ft-lbs 24 ft-lbs 30 ft-lbs 30 ft-lbs 15MLE 19.3 MLE 25MLE 25MLE 58.4 (56.5) 57.7 (56.4) 64.5 (63.3) 64.2 (63.2) 69.4 (65.6) 80.l (76.0) 89.2 (84.7) 87.6 (84.6) 82.8 (78.4) 92.5 (88.0) 106.6 (101.4) 104.8 (101.2) 87.7 (85.2) 101 (98.5) 124.4 (120.8) 122.4 (120.6) 13.4 (9.7) 12.0 (9.5) 11.6 (9.5) 11.1 (9.3) 34.8 (27.4) 33.9 (27.1) 34.l (27.4) 31.7 (27.2) 34.2 (27.1) 33.7 (27.2) 33.8 (27.2) 31.5 (27.0) 30.1 (26.4) 29.5 (26.3) 30.3 (26.5) 28.2 (26.3) 18.5 (17.5) 18 (17.5) 21.5 (21) 21.5 (21) 24 (22) 30 (28) 36 (33) 35 (33) 32 (29) 38 (35) 47.6 (44) 46.5 (44) 35 (33.5) 44(42) 61 (58) 59.5 (58) SA-517, Gr. F SA-517, Gr. E & F 3" 4" second set of fracture toughness values are listed in Case 6: S = Su/4 + 0.3 Sy residual stress parenthesis and were obtained using the ASME Section Case 7: S = Su/3.5 + 0.3 Sy residual stress XI, Article A-3000 method for K 1 determination. Case 8: S Su/3 + 0.3 Sy residual stress Increase in design stress ·from Suf4 to Suf3.5 Case 9: S Lesser of Su/2.4 + 0.3 Sy and% Sy+ 0.3 Sy (Table 15): Although this table shows an increase Case 10: S Su/3 + Sy (for SA-516 and SA-537) the required K1c with the increase in the design Case 11: stresses from Su/4 to Su/3.5, there is extensive S = Su/3 + 50 ksi residual stress (for SA-516 Gr. 70) experience with Section VIII, Division 2 vessels S = Su/3 + 71 ksi residual stress (for SA-517) (S = Su/3) with essentially the same toughness reSy = specified minimum yield strength (assumed to quirements as Division 1. be the same in base metal and in weld metal for Increase in design stress from Suf3 to lesser of these calculations). Su/2.4 and% Sy (Table 16): There is a substantial Cases 6-9 represent stress relieved welds, and Cases increase in the required Krc for SA-537 and SA-517 10-11 represent as-welded joints. The critical fl.aw when the design stress is increased from Su/3 to sizes in the SA-516 Gr. 70 weld were evaluated with Su/2.4. There is a lesser increase in the required Krc a residual welding stress equal to 38 ksi and 50 ksi, for SA-516 Gr. 70 due to the lower yield strength for comparison. The Su and Sy values for the SA-516, (S =%Sy governs). SA-537 and SA-517 steels are listed in Table 17. The Level 2 FAD procedure requires somewhat higher fracture toughness than Article A-3000 because of the stress ratio (Sr) in the Level 2 failure assessment diagTam. Welds. In the as-welded condition the residual SA-516, Gr 70 All stresses in some welds can be as high as the yield SA-537, Cl. 1 ts 2W' strength of the material. In PWHT structures the 2Yz" < t :5 4" SA-537, Cl. 2 t s residual stresses generally will not be reduced to zero. For this evaluation, the residual stresses in the as-welded joints are assumed to be as high as the yield strength of the weld for the SA-516 and SA-537 steels and 71 ksi (50 kg/mm2 ) for SA-517.39 In the Critical depths of surface flaws were evaluated PWHT condition, the residual stresses in the direcwith the Fracture Graphic® software using the tion of the weld are assumed to 30% of the yield Level 2 FAD procedure. The residual stress is asstrength of the weld for all steels. 26,36 The following sumed to act uniformly through the thickness. The calculations attempt to evaluate the critical fl.aw calculated critical flaw sizes (depth a) are given in sizes and the toughness needed for stress relieved Table 18 for aspect ratio all 0.15 (l/a 6%). welds to retain the same margin on brittle fracture Tables 19 and 20 give the required fracture tough(the same critical fl.aw size) for vessels designed to ness K 1c values (based on Level 2 FAD approach) for higher stresses as for vessels designed to the present stress relieved welds to retain the same critical Code allowable stresses. depth of surface flaws (with aspect ratio of all 0.15) Critical fl.aw sizes transverse to the direction oft}lfe with increased design stresses as with the present weld are evaluated for the following cases: Code design stresses and toughness requirements. SA-537, Cl.1 SA-537, Cl. 2 surface flaws with an aspect ratio of all = 0.15 if the design stresses are increased from Su/4 to Su/3.5 and from Su/3 to the lesser of Su/2.4 and % Sy, to assure the same critical fl.aw sizes as with the present Section VIII Division 1 and Division 2 design stresses 38 and toughness requirements. Two sets of values ar given for the required Kic and the Ch~rpy energy. The fast set of values designate the required fract toughness Kic obtained with the Fracture Graphi computer program and Level 2 FAD procedure. Th WRC Bulletin 435 SA-517, Gr. F SA-517, Gr. E 3" 4" 3" 4" lYz" 2" 3" 4" 0.392 0.422" 0.549" 0.572" 0.304" 0.416" 0.607" 0.635" 0.327" 0.416" 0.633" 0.663" 0.185" 0.251" Evaluation of Design Margins, Section VIII 0.183 0.185 0.233 0.234 0.116 0.157 0.202 0.202 0.120 0.151 0.211 0.211 0.095 0.128 39 From Su/4 + 0.3 Sy to Su/3.5 + 0.3 Sy (Table 19): Table 19 SA-516, Gr. 70 lY2 SA-537, Cl.1 2 3 4 1% 2 3 4 SA-537, Cl. 2 SA-517, Gr. F SA-517, Gr. E 3 4 1% 2 3 4 57.2 57.0 63.9 63.8 56 65.1 72.5 72.4 67 75 86.6 86.6 72.6 84 102.9 102.9 11.1 10.7 10.6 10.4 8.7 8.9 9.0 8.9 8.6 8.4 8.7 8.7 7.7 7.7 7.7 7.7 From Su/3 + 0.3 Sy to lesser of Su/2.4 + 0.3 Sy and% Sy + 0.3 Sy (Table 20): Table20 SA-516, Gr. 70 l\;2 SA-537, Cl.1 2 3 4 1% 2 3 4 SA-537, Cl. 2 lY2 SA-517, Gr. F 2 3 4 1% 2 3 4 SA-517, Gr. E 55.7 55.5 62.2 62.1 61.3 71.1 79.4 79.1 73 82 94.6 94.3 78.5 90.8 111.5 111.3 8.2 7.8 7.6 7.4 19 18.9 19.4 18.9 18.3 18.5 18.7 18.3 16.5 16.4 16.8 16.5 These calculations give an indication of critical fl.aw sizes with various stress levels and the percentage increase in the toughness requirements with increased design stresses to maintain the present margins against brittle fracture. There is a somewhat lesser percent increase in the toughness requirements for welded joints than for the base metal to provide the same margin against brittle fracture. This is because the percent increase in the total stress is less in welded joints. However, the critical fl.aw sizes in welded joints are smaller than those in base metal (when subject to design membrane stresses only) due to the additional welding residual stresses in welded joints. A decrease in the design margin requires a careful evaluation of notch toughness and NDE requirements to insure that there is adequate margin 40 against brittle fracture. Additional toughness and/or NDE requirements need to be imposed when the critical fl.aw sizes are smaller than the Code acceptance criteria (such as the maximum allowable defect sizes for RT examination in UW-51 of Section VIII, Division 1 and in Article I-5 of Division 2, the UT acceptance criteria in Code Case 2235, etc.) or where the combination of the applied stresses and the available toughness result in very small critical fl.aw sizes. A significant amount of satisfactory experience exists with vessels designed to Section VIII, Division 1 and Division 2 requirements having essentially the same toughness requirements, therefore, an increase in the toughness requirements due to increase in the design stress from Su/4 to Su/3.5 does not appear to be warranted. Additional consideration, however, should be given to the high strength quenched and tempered steels in Division 1, Part UHT and Division 2, TableAQT-1, regardless of the design stresses. These steels have high yield strength/ tensile strength ratios and high residual welding stresses in the as-welded and in the stress relieved condition. Both Division 1 and Division 2 already require essentially 100% volumetric and 100% SUI\ face examination of welds in the high strength quenched and tempered steels. Although there is good past performance record with the high strength steel pressure vessels designed to Division 1 and Division 2 requirements, additional consideration should be given to thicker plates (over about %" in as-welded condition), regardless of the design stress and for all thicknesses when the design stresses ar increased over Su/3. d. Yieldfl'ensile Strength Ratios Steels with different yield-to-tensile strength ra tios have different values of the slope of the stress~ strain curve in the inelastic range, local elongation. (elongation in the necking region) and uniform elon? gation (total elongation less local elongation). lri tension, both the local elongation and uniform elon" gation are important. Yield/tensile strength ratios. typically increase with increasing strength of steels High yield/tensile strength ratios limit the amoun of strain that can be safely tolerated at discontinui ties in pressure vessels of high strength steels (Par UHT in Section VIII, Division 1 and Article F -6 · Division 2). A significant experience already exist with pressure vessels constructed of high strengt steels (e.g. SA-517, SA-543). ASME Section VIII Divisions 1 and 2 already incorporate the necessar requirements for design, fabrication and NDE t insure a safe design. The following discussion is based mainly on th report by R. L. Brockenbrough & Associates. 30 Thi. report reviews the studies in Japan and in USA o the effect of yield/tensile strength ratios on stru tural behavior of high performance steels for bridg construction. The following is a brief summary o.. WRC Bulletin 435 some of the effects of yield/tensile strength ratios discussed in this report, which are significant to pressure vessels. Tension and Bending. Fig. 4 (Fig. 1.1 in Ref. 30) shows a reduced section of a tensile specimen (to simulate a bolted splice) and a typical stress-strain curve, where cry = yield strength, cru tensile strength, Ey = yield strain, Eu = strain at tensile strength, and Est = strain at initial strain hardening. The elongation of the member at ultimate load depends on the yield/tensile strength ratio. If the member is subjected to increasing tensile loads and the ratio is 1.0, the length of the member that yields approaches zero and the total elongation is limited. At smaller yield/tensile strength ratios a longer zone is able to reach yield strength while the minimum section is reaching the tensile strength, and thus the total elongation before rupture increases. Ref. 30 gives the following expression, developed by Kato, for maximum elongation omax in a tensile test specimen with length 2L: 20 15 e .§ (.,:) (.,:) .... G CJ > 0 c IO 0 ~~ g w G GG = 200mm I er-a;. ·-~\ 8\ \, •SS41 xSM50 oSM50Y OSM58 _J t= 10 mm r~ ~: 1 0 0.2 0.4 0.6 0 0.8 l.O Fig ..5-Effe~t of yie~d/tensile strength ratio on elongation capacity of tension specimen with a reduced section Eu Est fX A Ey .LL A +---Jn dx +-dx 0 1 0 Yo } YxAx Ao= area of the cross-section at mid-length, Ax= the cross-section area at point X which defines the limit of yielding (Axcry = Aocru or Axcry = Ao/Y, where Y = yield/tensile strength ratio). Figure 5 (Fig. 1.2 in Ref. 30) compares the above equation for a plate with a reduced section with test data for a 10 mm plate from four steels with yield/ tensile strength ratios from about 0.65 to 0.85. This figure shows the trend of decreasing elongation with increasing yield/tensile strength ratios, particularly at yield/tensile strength ratios over about 0.85. According to Kulak, 30 in a reduced section, x ::trs_ _ _ _ _:_ (1 I ' : I . I . c, 1 £u I • ' 0 . I ' ' I I £ cy Fig: ~:-Tension specimen with reduced section, and stress-strain def1nit1on behavior of a member depends on An/Ag and cry/cru. In some cases it may be desirable to have gross section yielding before fracture in the net section, to permit distortion before failure. This requirement would be satisfied theoretically with the following relationship: An/Ag::::: cry/cru. (If the Ancru < Agcry, the member deformation is limited). An = net cross-section area Ag = gross cross-section area of plate or tensile test specimen. cry= yield strength of the material. cru =tensile strength of the material. The above relationship, however, is not always practical since the actual properties, instead of the specified minimum values need to be considered. Also, the tensile stress through the net section at fracture is greater than the tensile strength of the material. Consequently, special details should be employed to prevent net section fracture before gross section yieik!ling in materials with high yield/tensile strength ratios. Fig. 6 (Fig. 1.13 in Ref. 30) shows the relation between yield/tensile strength ratio and uniform elongation. Total elongation in a tensile specimen consists of two parts, local elongation in the regions that necks and fractures, and uniform elongation in the remainder of the specimen. Uniform elongation is related to the yield/tensile strength ratio, tending to decrease with increasing yield/tensile strength ratios. According to Otani, 30 the most effective way to increase deformation capacity is to: Evaluation of Design Margins, Section VIII 41 30 ~,.,=Q.6(1-YRI 0 25 ~ "" 20 -~ ;;; co 15 ::..:; 0 0 f c: 0 ooo 'O 00 co 0 rte<:% oo 0 0 0 0 0 ~~00 00 0 c: 8, 0 ocSil 0 0 0 ocoo 0 -.; E 10 0 0 ~ '<:> gi 0 BQ;o o <Q 0 0 0 ~ 0 0 so 70 60 80 90 100 Yield ratio YR (O!o) Fig. 6-Relation between yield/tensile strength ratio and uniform elongation • Decrease the yield/tensile strength ratio for steels that have over about 10% uniform elongation, • Increase the uniform elongation in other cases. Winter 30 Dhalla and showed that the most important factor is the local ductility and that an elongation of 20% in a Vz" gage length (including the necking region) is sufficient to insure ductile behavior. Other researchers have shown that the elongation capacity increases with decreasing yield strength/tensile strength ratio. The yield/tensile strength ratio also has a significant effect on the rotational capacity of structural members with a moment gradient. The rotation capacity decreases rapidly with increasing yield/ tensile strength ratio, and decreases gradually with increasing yield point. If the yield strength is considerably less than tensile strength, the plastic region in a beam can extend over some length of the beam as bending moment at a critical section increases above a plastic moment by strain hardening. The spread of the plastic region contributes greatly to rotational capacity. If, however, the yield strength/ tensile strength ratio is 1.0, there can be no extension of the plastic region because the tension flange can rapidly reach its ultimate strain and rupture as the plastic moment is reached. Thus the inelastic rotational capacity of a structural member with a moment gradient approaches zero as the yield/ tensile strength ratio approaches 1.0. Pressure Vessels. The PVRC sponsored research at University ofKansas included burst testing oftest vessels made of three types of materials with different strain hardening components. 19 •20 The materials were SA-240 Type 304 stainless steel, SA-516 Grade 70 carbon steel and SA-517 Grade F quenched and tempered steel. These steels have strain hardening exponents of 0.585, 0.189, and 0.085, respectively. Burst tests were performed on test vessels with nozzles and fl.at heads and on test vessels with sharp 42 notches. These studies included the effects of stain hardening and the yield strength/tensile strength ratio on the bursting strength of the test vessels. These tests confirmed the validity of the modified Svensson eguation (PB= au Fey! ln W) for predicting burst pressure based on plastic instability and showed that, ignoring the effects of strain concentrations, pressure vessels of higher strength steels tend to burst at a higher percentage of their tensile strength than do lower strength steels. One set of tests at University of Kansas included tests of vessels with severe strain concentrations. The effect of a sharp undetected fl.aw was simulated with a notch 15" long in the axial direction of the vessel with a notch root radius of0.001" (0.025, mm). One set of vessels of each type of steel was tested with a notch depth of 0.18" (approximately 35% of wall thickness). Three additional A517 vessels were tested with a notch depth of 0.075", 0.125" and 0.18" (15%, 25% and 35% of the wall thickness). The report by Royer and Rolfe 20 also included the results of several sets of burst tests by others on test vessels of HY-140 steel with notches up to 40% of the wall thiclmess. The results of all the vessel tests with notches reviewed by Rolfe and Royer are included Fig. 7, which plots the burst pressure ratio PAIPB vs. notch depth in percent of wall thickness. (PA = actual burst pressure, PB = theoretical burst pressure based on plastic instability). The results of these tests led the investigators to believe that the percent reduction in burst pressure from the Svensson prediction is directly proportional to notch depth as percent of wall thiclmess. For the A 517 steel the reduction in burst pressure exceeded the reduction in wall thiclmess for the 35% notch depth. At some point beyond a notch depth of 25%, the burst failure could no longer be predicted accurately from the Svenssen equation by assuming reduction in burst pressure as a percentage of the wall thickness. For notch depth greater than about 25% of wall thiclmess the reduction in burst pressure in high strength steels (with high yield/tensile strength ratios) appears to be dependent upon ductil- ity, toughness, and possibly the strength-to-toughness ratio of the steel. Langer 17 •30 derived expressions for strain concentration factors (K.) in a tapered, axially loaded bar and at the end of a cantilevered beam in terms of the strain hardening exponent, n. The strain concentration was defined as the actual peak strain divided by peak strain calculated for completely elastic behavior, based on the assumption that the maximum deflections are the same. Fig. 8 shows that the strain 1-t ~!---·-_-_- y (b) Cantilever beam. (a) Tapered bar. 4o.--~~~~~~~~~~~~~~~~--. 30 c 304SS • A516 A517 - lst Heat • A517 - 2nd Heat 20 111 o HY-lMXTl A AlOOB 10 ~o..a:i o" 0.8 j:: < ci:: 8 A •8 OA ~ ~ 0. 6 A ""'• ci:: 0.. ,_ VI ci:: :::l "Xw 6 0 4 A A 3 Ill 0.4 2 al 0.2 0 I 0 2 0 3 0 4 0 5 0.6 0.7 0 8 0.9 1.0 n 80 60 NOTCH DEPTH. '1o of WALL THICKNESS 20 25 Fig. ?-Effect of notch depth on burst pressure WRC Bulletin 435 100 (c) Strain concentration factors. Fig. 8-Relation of strain hardening exponent n to strain concentration factors K, Evaluation of Design Margins, Section VIII 43 concentration increases as the strain hardening exponent n decreases in axially loaded bar with tapered notch and in cantilever beam under bending stresses. Material with low strain-hardening exponent, such as high strength steel with high yield/tensile strength ratio, will tend to have a higher strain concentration in the presence of a notch. In steels with higher strain hardening exponents, the material surrounding the notch tends to control the strain in the plastic zone. Fatigue. The report by R. L. Brockenbrough30 includes a summary on large scale welded HSLA-80 steel fabricated sections for ship construction. Details included longitudinal fillet welds, transverse groove welds, bulkhead attachment details, etc. The lower confidence limits of the S-N curves were not significantly affected by mean stress, and were not significantly different than those for similar weld details in structural steels. It was, therefore, concluded that fatigue strength of welded details in air can be considered to be independent of the strength and type of steel. e. Fabrication and NDE Although difficult to quantify the Code requirements for fabrication and NDE with respect to margin on safety, they are essential to providing the necessary design margin in the structure. Design details, fabrication procedures and NDE become even more important with lower design margins. Parts UCS and UHT in Section VIII, Division 1 and Part AF in Section VIII, Division 2 include specific requirements to minimize the risk of degradation of materials properties during fabrication, limitations on fabrication defects, etc., some of which are discussed below. Forming. Excessive forming strains (without subsequent heat treatment) can degrade strength and/or toughness of steels. Section VIII, Division 1 includes the following limitations on the extreme fiber elongation during cold forming without subsequent PWHT: a) UCS-79 limits the maximum elongation to 5% when: the vessel contains lethal substances the material requires impact testing thickness before forming exceeds %" the reduction of thickness by cold forming is more than 10% 5) the temperature of the material during forming is in the range of250°F to 900°F. 1) 2) 3) 4) The extreme fiber elongation can be as great as 40% for P-No. 1Group1 and 2 materials if none of the 5 conditions above exist. b) UHT-79 (high strength quenched and tempered steels) limits the maximum elongation to 5% in all cases. 44 Division 2 has the following requirements: a) AF-111. All materials for shells sections and heads shall be fored to the required shape by any process that will not unduly impair the mechanical properties of the material. However, no testing is required after forming. b) AF-605 limits the maximum forming strain to 5% max. for the quenched and tempered steels in TableAF-630.1. Some non-USA pressure vessel codes have addi.: tional requirements. BS 5500 requires stress reliev-. ing of all cold formed heads unless the manufacturer demonstrates that the cold formed properties are adequate and the material properties are not significantly altered. (This is because substantial forming strains may exist in certain dished and flanged heads which cannot be calculated by the formulas given for bending strains). For cold formed cylindrical parts, BS 5500 restricts the minimum inside radius to 10 t for C and C-Mn steels and to 18 t for other ferritic steels. Tolerances. The out-of-roundness tolerances are similar in ASME Section VIII and in some of the European pressure vessel codes. Specific require~ ments for maximum deviation from a circular form are also provided in both Section VIII, Division 1 an Division 2. Section VIII, Division 1, UG-79(b) requires th~ adjoining edges oflongitudinaljoints to be shaped to proper curvature to avoid objectionable flat spo along the completed joints. Both Division 1 an Division 2 include specific limitations on the maxi2 mum offset in welded joints. Neither Division 1 nor Division 2 include specific requirements for maxi~ mum peaking at welded joints, whereas detaileg requirements are included in the draft of the CE~ pressure vessel standard. Local deviations (e.g. a longitudinal joints in cylinders, cones, dished head spherical shells) can increase local stresses whic may need further evaluation under cyclic loadin conditions, particularly when considerations is give to increased allowable design stresses. NDE. ASME Section VIII, Division 1 permit either 100% RT, spot RT or no RT examined vessel The joint efficiencies are assigned accordingly, i. 100% for full RT, 85% for spot RT of double bu welded joints, etc., depending on the type of welde joint (Table UW-12). No specific examination requir ments are given in Section VIII, Division 1 for MT o PT examination of nozzle welds for C, C-Mn, or lo alloy steel materials in Part UCS. Higher design stresses necessitate additional r quirements for NDE of welded joints. This is r fleeted in ASME Section VIII, Division 2 requir ments, where essentially all welded joints must b non-destructively examined. Division 2 requir 100% RT of all butt welds and 100% surface examin tion of all nozzle attachment and other attachme welds. Additional MT or PT examination requir WRC Bulletin 435 1nents are specified in Division 2 for radially disposed cut edges in vessel walls for nozzles to be attached by fillet welds, partial penetration welds, temporary attachment weld locations, as well as PT examination of austenitic chromium nickel steel welds exceeding%" thickness. Fracture mechanics calculations show that critical fl.aw sizes decrease as the design stresses, yield strength and thickness increase. The calculations in Section E.2.c of this report resulted in relatively small critical flaw sizes for the high strength quenched and tempered steel SA-517 in the aswelded condition. Stress relieving of welds in the high strength Q & T steels is done at somewhat lower temperatures than other carbon and low alloy steels, and significant residual stresses may remain in welded joints of these steels after PWHT. (See Table UHT-56 in Section VIII, Division 1 and Table AF-630.1 in Division 2 for PWHT requirements.) The Section VIII toughness requirements and/or NDE requirements, therefore, need to be re-examined for these steels when consideration is being given to reducing the present design margins. (Several European pressure vessel codes require additional surface examination of welded joints, including butt welds, with MT or PT methods for the higher strength and more crack sensitive materials and for higher risk category vessels.) While ASME Section VIII, Division 2 requires essentially 100% examination of nozzle attachment welds, Division 1 vessels presently have no specific requirements for nondestructive examination of nozzle welds. NDE requirements for nozzle welds need to be included also in Division 1 when consideration is given to higher allowable design stresses. Welded Joints. Various types ofweldedjoints are permitted in the Code for Division 1 vessels. Only Type 1 and Type 2 (for Category B joints only) are permitted for butt joints in Division 2 vessels, and only Type 1 butt joints are permitted for the high strength materials in TableAQT-1. Type I butt joints are double welded full penetration shell joints, and Type 2 are single welded butt joints with backing strips which remain in place. British Standard 5500 also permits single-welded (with or without backing strips) or double-welded butt joints for main seams in vessels; however, single-welded butt welded joints are limited to 20 mm thickness, or less, depending on joint detail. Also, permanent backing strips must not be used for Construction Category 1 and 2. 5 This should be an additional consideration for Division 2 vessels subject to higher design stresses. PWHT. ASME Section VIII, Divisions 1 and 2 require PWHT of welded joints for thicknesses exceeding Viz" in carbon and C-Mn steel vessels provided welded joints over 1 Yi" are preheated to 200°F. European pressure vessel codes generally require PWHT at lesser thicknesses (30 mm in the ADMerkblatt Code and 35 mm in BS 5500 and in the Draft CEN Standard), but most European pressure vessel codes also have provisions for increasing the thickness limits for PWHT based on additional toughness, fracture mechanics analysis and NDE. (Although the benefits of PWHT are well known, the authors do not think it necessary to have more restrictive PWHT requirements in Section VIII as a justification for higher allowable design stresses). 3. Review of Vessel Failures A general discussion on pressure vessel failures was included in the Phase 1 Report of this project. An attempt was made by the authors to gather further information on pressure vessel failures related to specific design codes (ASME Section VIII, Divisions 1 and 2, BS 5500 and AD-Merkblatt). Although published literature contains numerous articles on pressure vessel and piping failures, 22 no comprehensive summary exists on failures grouped with respect to different codes or standards, nor vessels designed to different allowable design stresses. The information available to the authors indicate that pressure vessel failures generally fall into the following categories: 1. Failures from inadequate design or poor design details for the specified design conditions (e.g. fatigue). 2. Failures due to poor notch toughness of materials, fabrication defects, welded repairs, etc.). 3. Failures caused by service related degradation of materials and cracking (corrosion, stress corrosion cracking, etc.). 4. Process or operation related failures of pressure vessels (overheating, explosions, etc.) 5. Failures from a combination of the causes listed above. The majority of failures can be attributed to poor design details or improper fabrication practices for the materials used, poor notch toughness (particularly during overload testing), service related degradation, and operation related problems. The incidence of catastrophic failures due to design faults is low. No failures are known to the authors which can be attributed directly to the design rules in the recent editions of the codes and standards reviewed as part of this report. 4. Summary and Conclusions The authors have reviewed in Section E of this report the various failure modes considered inASME Section VIII, Division 2 and certain factors which influence the design margins. Factors, such as design details, yield-to-tensile strength ratios, ductility, fabrication and non-destructive examination are difficult to quantify with respect to margins of safety, and the authors have provided a general discussion on the significance of these factors. In absence of other failure modes (such as fatigue Evaluation of Design Margins, Section VIII 45 and brittle fracture), the design of Section VIII, Division 2 vessels is generally governed by the bursting mode of failure when the yield-to-tensile strength ratio is more than 0.5. The vessel tests at University of Kansas on test vessels of materials with three different strain hardening exponents showed that stainless steel vessels (which have a high strain hardening exponent) designed to Division 2 stresses can have a true margin of safety less than 3.0 against the bursting mode of failure. The A 517 vessels with moderate strain concentrations had a true margin of safety over 3.0 against this failure mode. Fracture mechanics calculations were performed on several materials using the BS PD 6493:1991 Level 2 FAD procedure to evaluate critical flaw sizes based on the present Code design stresses and toughness requirements and for higher design stresses (S = Su/3.5, and the lesser of % Sy and Su/2.4) with the present Code toughness requirements. Two sets of calculations were performed to evaluate the toughness needed to insure the same margin of safety on brittle fracture with the increased design stresses. One set was based on the present Code allowable stresses and with increased allowable stresses in the base metal. Another set was performed on welded joints assuming that the remaining residual stresses after PWHT are about 30% of yield strength of the weld. The first set of calculations resulted in 16-23.3% increase in the required fracture toughness (Krc) for increasing the design stress from Su/4 to Su/3.5 for all the steels and thicknesses considered in this evaluation. The increase in Krc was 26.3-27.4% for SA-537 and SA-517 for increasing the design stress from Su/3 to the lesser of Su/2.4 and% Sy and 9.3-9.7% for SA 516, Gr. 70 (because of the lower yield strength of this material). The second set of calculations resulted in a somewhat lesser increase in Krc for the welded joints since the relative increase in the total stresses (primary plus secondary stresses) was lower. No specific increase in the required fracture toughness was recommended by the authors for increasing the design stresses from Su/4 to Su/3.5 due to the extensive successful experience with ASME Section VIII, Division 2 vessels which have essentially the same toughness requirements as Division 1 vessels. However, a further review of the Code requirements is suggested for the high strength Q & T steels in Table UHT-23 of Division 1 and Table AQT-1 of Division 2. Most of these steels have a high yield strength-to-tensile strength ratios, less capacity for strain concentrations at notches and discontinuities than other materials and can have high residual stresses in welded joints. Therefore, these materials need additional considerations for notch toughness and/or NDE, regardless of design stresses. Other factors discussed in Section E which can have an effect on the margin of safety are cold forming strains, fabrication tolerances, non-destructive examination, welded joint details and PWHT. 46 All these factors need to be considered when increasing the allowable design stresses, and additional Code requirements need to be imposed (with the exception of PWHT) to insure that the present safety margins are not jeopardized. F. Comparison of ASME Section VIII, Divisions 1 and 2 with Other Codes and Standards This part of the Phase 2 study on margins of safety in ASME Code, Section VIII includes a review several international pressure vessel codes and standards, which use lower design margins than ASME Code, Section VIII, Divisions 1 and 2, and comparison of some of the essential requirements in these international codes and standards with those in ASME Section VIII, Divisions 1 and 2. The true margin of safety depends on all of the essential requirements in a particular code or standard: the basic design rules, design details, materials and toughness requirements, ductility, fabrication requirements, NDE, overload testing, overpressure protection, quality control, and a number of other factors. It is not possible to include all considerations affecting the safety margins in this study, therefore only those factors deemed more significant to the safety of pressure vessels are addressed. This study addresses the following codes and standards: 1. 2. 3. 4. ASME Section VIII, Division 1 ASME Section VIII, Division 2 British Standard 5500 The German Pressure Vessel Code (AD Merkblatt Code) 5. The Netherlands Rules for Pressure (The Stoomwezen Code) 6. CEN Draft Standard for Unfired Pressure sels. This part of the Phase 2 study compares those requirements in these codes and standards which have a significant effect on margins of safety of welded pressure vessels. The following Code requirements are reviewed and compared in this report: 1. 2. 3. 4. 5. 6. 7. 8. 9. Design margins Design rules Design details Service considerations Materials requirements Toughness requirements Fabrication requirements Overpressure testing Quality control requirements A more detailed summary of each of the non-USA codes and standards listed above is included in Appendices 1 through 4 of the Final Report, PVRC Project No. 97-2.40 The codes and standards used in this report were the latest available to the authors at the time of the report and may not all be the latest editions issued. WRC Bulletin 435 1. Design Margins a. General The design margins, although not a sole measure of the safety of a pressure vessel, are the most significant factor in evaluating comparative safety. Design margins, reflecting an "ignorance factor" have historically been reduced as the understanding of behavior, experience base, and methods of design and analysis have improved. The design margins on yield strength have universally remained at 1.5 (except for austenitic stainless steels and other strain hardening materials, lower factors have been used). The definition of yield strength varies somewhat from one code to another, but they are not significantly different. The margin on ultimate tensile strength, however, has changed over the years and varies a great deal from one code to another. This margin for the ASME Code, a pioneer code for pressure vessel construction, was 5 for many years. In 1940's the confidence with these rules and the urgent needs of the period were the justification for a reduction of this factor to 4. With the advent of design by analysis, most codes in the 1960's required more analysis and quality control but reduced this design margin to 3. One such code was Division 2 of ASME Section VIII. Another ASME Code using a design margin of 3 on ultimate tensile strength is the B31.3 piping code. B31.3 has about 50 years of satisfactory experience with the use of that code. British Standard BS 5500 allows a design margin of 2.35 on the specified minimum ultimate tensile strength for Construction Category 1 and 2 vessels. 5 This factor, compared to that of ASME codes is even more liberal than it appears. While the design margin on yield strength is applied to the value at design temperature, the factor of 2.35 is applied to ultimate strength at room temperature, regardless of design temperature. (This factor for austenitjc stainless steels is 2.5.). For Category 3 vessels {if carbon steel this factor is 5. However, this factor reflects the joint efficiency for these vessels which get no volumetric examination of the seams. This fact will make this factor more liberal than Section VIII, Division 1 (VIII-1), for joints with no radiographic examination (RT). For Construction Category 3 austenitic steel vessels, there is a specified value to be used, regardless of the type of austenitic steel. One factor which slightly contributes to design margins of BS 5500 is the fact that material thickness undertolerance is to be accounted for, whereas such undertolerances are disregarded in the ASME Codes and assumed to be covered by design margins. Another major difference is the fact that BS-5500 provides a great deal of flexibility in rules, even as compared with Section VIII, Division 2 (VIII-2). Overall, the design margins on ultimate tensile strength in BS-5500 are considerably more liberal than those ofVIII-1 and VIII-2. In the German AD-Merkblatt (ADM) pressure vessel code the design margin of 1.5 on yield strength is generally applicable to rolled and forged steels. 6 Factors on ultimate tensile strength are specified only for cast steels. Another factor contributing to these rules being liberal is the fact that the stresses due to dead loads, supports, wind and snow do not have to be accounted for, as long as these stresses do not add up to more than 5% of the material stress allowable. Overall, the design margins of this code are considerably more liberal than those of VIII-1 and VIII-2, for those materials with stress allowables based on ultimate tensile strength in the ASMECodes. The Dutch Stoomwezen pressure vessel code accounts for the ductility of materials in establishing design margins. 7 For materials with greater than 10% elongation, the design factor on ultimate tensile strength is 2.27, and with avoiding details that result in high peak stresses, this factor may be reduced to 2.0. These factors are applied to ultimate tensile strength at room temperature, regardless of design temperature. However, for materials with elongations of 10% or less, a factor of 4.0 applied to ultimate tensile strength at design temperature is specified. This code also requires that material undertolerances be accounted for. For ferritic steels with high yield strength or yield to tensile ratios greater than 0.67, certain requirements on details and fatigue exemption calculations are applicable. A feature of this code is the fact that allowable stresses are modified for certain components. A direct comparison with ASME Codes is not practicable. Overall the design margins on ultimate tensile strength of this code are considerably more liberal than those of VIII-1 and VIII-2, for ductile materials. For a few materials with low ductility, these rules may not be more liberal. In the draft of proposed CEN unfired pressure vessel standard, the basic ultimate tensile strength design factor for ferritic steels is 2.4. 8 However, this factor may be reduced to 1.875 for the Alternative Design Basis, which requires certain controls on details, examination, and quality control. (Alternative Design Basis was included in the July 1995 draft of the CEN/TC 54 Standard for Unfired Pressure Vessels but not in the May 1998 draft). Austenitic steels with elongation less than 30% are treated the same as ferritic materials. For austenitic steels with 30% or elongation, there is no limit based on ultimate tensile strength. It should be pointed out that the rules of this standard are applicable only to those materials with a minimum of 14% elongation and a minimum impact value of 20 ft-lb at 20°C (68°F). In this standard, the material specification undertolerances are to be accounted for. A unique feature of this standard is the fact that stress allowable tables are not provided. Another unique feature is that for low probability "exceptional design conditions" lower design margins may be used by agreement. In general, this proposed standard Evaluation of Design Margins, Section VIII 47 contains the most recent requirements in Europe but also the most liberal design margins. In the creep range, the allowable stresses in the various documents are based on creep and stress rupture. Section VIII, Division 2 does not allow design of vessels temperatures in the creep range. For Section VIII-1, the stress. allowable stresses in the creep range are based on the lowest of the following: • 100% of the average stress to produce a creep rate of0.01% per 1000 hours. • 67% of the average stress to cause rupture at the end of 100,000 hours. • 80% of the minimum stress to cause rupture at the end of 100,000 hours. For welded pipe or tube, the allowable stress is 85% of that calculated on the above basis (structural quality ferrous materials are not allowed in the creep range). In the BS-5500, for temperatures in the creep range, the allowable stress is based on the mean value of the stress required to produce rupture in a specified time divided by a factor of 1.3. This is about 77% of the mean value compared with 67% of average value for VIII-1. No limit on creep is specified. This is generally more liberal than rules of VIII-1, ifthe design life is 100,000 hours. For certain materials, BS-5500 allows design life ofup to 250,000 hours with the stress basis being the same. For an VIII-1 vessel that may operate up to 250,000 hours (VIII-1 does not limit design life), the rules of VIII-1 would be very non-conservative relative to BS-5500. If the rules ofVIII-1 are made more liberal, a limit on life of vessels operating in the creep range should be considered. In the ADM Code, the stress allowable in the creep range is lowest of the following: • The average stress to cause rupture in 100,000 hours • The minimum yield strength • 1% strain based on proof test The rupture criteria is more liberal than that of VIII-1. The other two factors are unique to this code and are not expected to control the allowable very often. A direct comparison with criteria of VIII-1 is difficult. In the Stoomwezen code, the allowables stresses in the creep range are based on the lower of average stress which causes 1% creep in 100,000 hours or stress which produces rupture in 100,000 hours. This appears the same as the creep criteria ofVIII-1, but is in reality more conservative. The rate of creep increases with time so that .01%in1000 hours results in more than 1% in 100,000 hours. The stress rupture criteria of Stoomwezen is, however, more liberal. In the draft CEN pressure vessel standard no basis for stress allowables in the creep range has yet · been specified. 48 The following is a brief summary of the bases for design margins at temperatures below the creep range, as listed in each of the codes and standards addressed in this report: The allowable stresses for bolting materials are the lower of Yi Su or% Sy, except for quenched and tempered bolts (Section II, Part D, 2-120 ofAppendix 2). c. ASME, Section VIII, Division 2 b. ASME, Section VIII, Division 1 The basis for establishing allowable design stress values are summarized in ASME Section II, Part D, Appendix 1. At temperatures below the creep range, the maximum allowable stress value S is the lowest of the following: • • • • Su/4 at room temperature Su/4 at temperature 2Sy/3 at room temperature 2Sy/3 at temperature Sy = specified minimum yield strength Su = specified minimum tensile strength For austenitic steels and for certain non-ferrous alloys, two sets of allowable stress values are provided. The alternative allowable stress values exceed the 2Sy/3, but do not exceed the 90% of Sy at temperature, and are not recommended for use where higher deformations may be objectionable. In the application of these criteria, the yield strength at temperature is considered to be the specified minimum yield strength at room temperature multiplied by the ratio of the average temperature dependent trend curve value of yield strength to the room temperature yield strength. The tensile strength at temperature is considered to be 1.1 times the specified minimum tensile strength at room temperature multiplied by the ratio of the average temperature dependent trend curve value of tensile strength to the room temperature tensile strength. It should be noted that elevated temperature tensile testing of each heat of material used in Section VIII construction is not required. (Such testing is required by most European Codes.) It should also be noted that the trend curves are derived from ratios of elevated temperature tensile test data to room temperature tensile test data for each heat of material (based on average properties). The curves are then the statistical best fits of these ratio curves and anchored to the minimum specified properties to establish the tensile and yield strength values to which the allowable stress factors are applied. The elevated temperature tensile and yield strength values thus established are not the minimums, therefore, there is about 50% probability that the actual elevated temperature tensile and yield strength values might be lower than the values used to establish the allowable stresses. This fact takes away some of the apparent conservatism indicated by the factors used. The allowable stresses for structural materials used for pressure parts, bars and shapes used for shell stiffeners, or stay bolts are multiplied by a quality factor of 0.92. WRC Bulletin 435 Section VIII, Division 2 uses the Tresca, or maximum shear stress criteria to establish the allowable stress intensities. The stress intensity is the maximum allowable value of the stress difference. The following are the basic stress intensity, Sm limits in Division 2: Combination of Stresses General primary membrane Local primary membrane (PL) Primary membrane + bending (PL + Pb) Primary + secondary (PL+ Pb+ Q) Stress intensity, Sm limits k Sm :s k(2/3 Sy, or Su/3) 1.5 kSm 1.5 kSm (a) Material with specified elevated temperature strength values: (1) Up to and including 50°C (122°F): fE = Rjl.5 or Rm/2.35, whichever is lower Rm specified minimum tensile strength at room temperature Re specified minimum yield strength at room temperature (2) 150°C (302°F) and above: fE RecT/1.5 or Rm/2.35, whichever is lower Re(Tl = 0.2% yield strength (Rpo. 2 ) for carbon, C-Mn and low alloy steels (l.0% yield strength for austenitic stainless steels) specified for the material at temperature T (tested in accordance with BS 3688). (3) Between 50°C and 150°C: fE is based on interpolation between the values at 50°C and 150°C. (b) Material without specific elevated temperature strength values: 3Sm The following is a summary of design margins for the various load combinations (See Section II, Part D, Appendix 2 for detailed explanation): • Su/Pm :5 3/k • Su/PL :s 2/k • Su/(PL + Pb) :s 2/k (1) Up to and including 50°C: fE Rjl.5 or Rm/2.35, whichever is lower (2) 150°C and above: fE = Rjl.6 or Rm/2.35, whichever is lower (3) Between 50°C and 150°C: fE is based on interpolation between the values at the two temperatures. Austenitic Stainless Steels: Su specified minimum tensile strength Sy = specified minimum yield strength Pm = General primary membrane stress PL = Local primary membrane stress Pb = Primary bending stress k = stress intensity factor for various loading conditions The following table lists the stress intensity k factors from Section VIII, Division 2, Table AD-150.1 flr various load combinations: Load Combination Design pressure + dead load (DL) Design pressure + DL + wind load Design pressure + DL + earthquake load Design pressure + DL + wave action k Factor 1.0 1.2 1.2 1.2 The use of the trend curves is the same as for Section VIII, Division 1. d. BS 5500. The design strength values for various materials are established in accordance with Appendix Kin BS 5500. Basis for time independent design strength, fE (BS 5500, Appendix K): Carbon, Carbon-Manganese and Low Alloy Steels: (a) Material with specified elevated temperature strength values: (1) Up to and including 50°C: fE = Rjl.5 or Rj2.5, whichever is lower (2) 150°C and above: RecT/1.35 or Rm/2.5, whichever is lower. (3) Between 50°C and 150°C: Interpolate between the values at the two temperatures. Aluminum and Aluminum Alloys (annealed materials for welded construction): (a) Time-independent Design Strength: fE = Rpo. 2/ 1.5 Design Stresses for the Various Construction Categories in BS 5500: Categoriet1Pl and 2: The nominal design strength values in Table 2.3 (subject to the additional requirements in BS 5500). Category3: (a) Carbon and carbon-manganese steels fE :5 Rm/5 (b) Austenitic steel fE :s 120 N/mm 2 or 120 (450/(400 + t)), whichever is smaller. t = design temperature, °C (300°C max. for Construction Category 3 vessels). Evaluation of Design Margins, Section VIII 49 For materials with specified yield stress (1 % proof stress) :5 230 N/mm 2 , the allowable design stress so calculated shall be multiplied by 0.8. e. AD Merkblatt The allowable design stress is given by K/S, where the K value is the specified strength value at the design temperature (generally the specified minimum yield point) and S is the safety factor. (See AD-Merkblatt B 0 and B 1). The safety factors Sare listed in Table 21. f. Stoomwezen At temperatures below creep range, the nominal design stress shall not exceed the smaller of the following values: Materials with elongation As > 10%: The design stress f1 shall not exceed the lower of the following values: Materials with elongation As < 10%: • f1 :s 0.25 Rm • fi :5 0.25 RmCvnJ RmCvm) tensile strength at temperature Vm g. CEN The CEN draft Unfired Pressure Vessels Standard lists the following bases for allowable design stresses: Design Stresses for Other Than Austenitic Steel for Pressure Parts Other Than Bolts: • The yield strength or 0.2% proof strength at the design temperature, divided by a safety factor 1.5. • The specified minimum tensile strength at 20°C (68°F) divided by a safety factor of2.4. Alternative Design Stresses: The Alternative Design Basis was included in the July 1995 draft of CEN/TC54 Standard for Unfired Pressure Vessels, but not in the May 1998 draft of this standard. The allowable stress values were not to exceed the smaller of the following: • The yield strength or 0.2% proof strength at the design temperature, divided by a safety factor of 1.5. • The minimum tensile strength at 20°C divided by a safety factor 1.875. 50 • 1% proof strength at the design tern perature divided by safety factor SF 1.5. (2) With minimum rupture elongation not less than 35%: The nominal design stress shall not exceed the greater of the following two conditions: • The 1% proof strength at the design temperature divided by SF = 1.5. • The smaller of the following: a. The minimum tensile strength at the design temperature divided by SF 3.0. b. The 1% proof strength at the temperature divided by SF = 1.2. Testing Condition: • fi:::;:; % ReCvm) or • f1 :::;:; 0.44 Rm ReCvm) =yield stress at temperature Vm Rm = tensile strength at 20°C (68°F) Rolled and forged steel and Aluminum and its alloys Cast steel Cast iron with spheroidal graphite Gray cast iron, Copper and its alloys Design Stresses for Austenitic Steels: (1) With minimum rupture elongation not less than30%: (1) Steels other than austenitic and austenitic steels with rupture elongation not less than 30%: 1.0% proof strength at test temperature divided by safety factor 1.05. (2) Austenitic steels with minimum rupture elongation not less than 35%: The nominal test stress shall not exceed the following: • The 1.0% proof stress at test temperature divided by safety factor 1.05. • The minimum tensile strength at test temperature divided by safety factor 2. Design Stresses for Steel Castings: The nominal design stress shall not exceed the smaller of the following: • The yield strength or 0.2% proof strength design temperature divided by safety factor 1.9. • The minimum tensile strength at 20°C divided by safety factor 3.0. 2. Design Rules a. Design for Internal Pressure The basic formulas for design of cylindrical, spherical and conical shells are almost the same in all: various codes and standards. The allowable stresses and joint efficiencies, however, result in widely different thickness requirements. The rules of Section VIII-1 result in much thicker shells than do the rules of other documents considered here. The rules of 1.5 2.0 2.4-5.0, depending the type of cast iron See Table 3, AD-Merkblatt B 0 WRC Bulletin 435 1.1 1.5 1.2-2.5, depending the type of cast iron See Table 3, AD-Merkblatt B 0 Section VIII-2 also result in thicker shells than all the European documents studied. The European documents mostly require that materiel specifications undertolerance be accounted for in the calculated thickness. They also have more explicit requirements regarding adequate allowance for forming and for corrosion (by adding forming and corrosion allowances to the calculated thickness). The rules for design of cone-cylinder junctions vary considerably. The European documents provide rules based mostly on analysis and consider loads other than pressure. Their rules are probably more accurate but their design margins are less. The boundary between thick and thin shells is also defined differently in various documents. European codes offer more guidance in calculation and evaluation of loads other than pressure (such as wind and seismic). The increase in allowable stresses for these environmental loads is similar in various documents. Another difference between European Codes and VIII-1 is the fact that VIII-1 does not have stress limits for test condition. Section VIII-2 and the European codes and standards referred to in this report do have such limits. b. Design of Dished Heads The rules for design of hemispherical heads are generally the same as those for spherical shells, in each document. The methods for design of ellipsoidal and torispherical heads, however, vary significantly. The German Code allows only 2 types of such heads. The basis for the design rules are not clear, but safety factors are applied to the stresses in the dome. Openings are not allowed in the knuckle area, without a significant penalty factor. Straight skirts of at least 3.5 times the shell thickness are required. The proposed CEN rules are mostly based on recommendations ofWRC Bulletin 364 (It should be pointed out that the authors of that Bulletin hare proposed more recent methods which are considerably more liberal. These proposed rules are being considered for adoption by the ASME Code Committee). The existing ASME rules as well as those of BS5500 are mostly based on the method proposed by Shields and Drucker (VIII-1 rules have their roots in some very old experimental data). A number of factors have been applied in various sets of rules, which makes a direct comparison impracticable. However, it is safe to say that VIII-1 and VIII-2 rules in this area are very conservative, except for thinner heads. Buckling of the knuckle is not addressed in the ASME Codes, and as a result an upper limit on D/T of 500 is specified. Some of the European codes do address buckling and allow thinner heads. The rules for spherically dished covers are somewhat different in various codes. The formulas for the spherical dome is mostly the same. The formulas for flange design are different. With the higher design margins, the rules ofVIII-1 are mostly more conservative than other documents. c. Design of Flat Covers The basic formula for design of fiat covers is the same in all codes. The allowable stresses and "C" factors representing boundary conditions are different. Unlike the ASME codes, most European codes do not apply a joint efficiency to the allowable stress. Since the stresses are primarily bending, applying a joint efficiency intended for joints subjected to membrane tension is too conservative. The diameter used for calculating bending stresses is different in some codes from the shell diameter used by ASME codes. Even though some of the "C" factors in the European codes are larger than the corresponding values in VIII-1 and VIII-2, the higher design margins of the ASME codes will mostly result in greater thicknesses, specially for higher strength materials or those with high yield to ultimate ratio. For materials with yield strength controlling the stress allowable, ASME rules, specially VIII-2, could result in lower thicknesses than some European codes. d. Design of Flanges Design rules for bolted flange connections, in BS-5500 and ASME Codes, are based on "Taylor Forge Method." The same gasket factors are specified in these documents. ADM rules are also based on this method, but the gasket factors, details of application, and weld detail requirements are different. This German Code puts a limit on allowable flange rotations. The method of Stoomwezen has a more analytical basis, and is quite different from other codes. This code also has a limit on flange rotation. In addition to the basic method, this code allows a number of other design and analysis alternatives, including the rules of VIII-1. The proposed rules in the CEN standard are also based on Taylor Forge method, with a number of differences in application. This standard provides for a more analytical method in Annex B. Due to many differences, it would be difficult to compare all these rules, but in general all the various sets of rules result in very comparable flange designs (it should be pointed out that the ASME Code is considering a totally new method of flange design). All of the documents studied, allow the use of standard flanges and, generally, such standard flanges are lighter than those which would result from the use ,?f design rules. e. Design ;} Openings The German ADM code includes two methods for reinforcement of openings. One method, based on pressure area, is similar to the method ofVIII-1; but there are a number of differences in details of application. The second method based on a "weakening" factor is a unique method. The definition of an isolated opening is different from that ofVIII-1. The rules for multiple openings, based on pressure area, are not as detailed but are similar to those ofVIII-1. The rules in BS-5500 are completely different from area replacement rules ofVIII-1. However, the use of Evaluation of Design Margins, Section VIII 51 a simplified area replacement method i~ allowed as an alternative. The basic rules are derived from a shakedown analysis and consider some minor piping loads. The rules of Stoomwezen are totally different from those of VIII-1. Definition of a single opening is also different. Rules for multiple openings are quite detailed and different. The applicability of these rul~s is limited to dJD ratio up to 0.7. It can only be said that this method is more analytical and more detailed than that ofVIII-1. The rules of CEN draft are still different from other codes and are based on a mix of limit analysis and pressure area approaches. Use of .reinfor?in.g pads is limited. The rules include a maximum h~mt on the ratio of nozzle neck thickness to shell thickness which has no parallel in any other code. N~ne of the codes studied have requirements for limiting bending stresses at the junction of large openings in cylindrical shells'. except _YIII-1. The rules ofVIII-1 have for a long time reqmred that for large openings % of stiffening area be p:a~e~ close to the junction. A second requirement for hm1tmg bending stresses, based on McBride and Jacobs pape:·, was added (Appendix I-7(b)), but it appears t~at this requirement may result in overly conservativ~ design for certain cases and may be deleted or mod~fied. The CEN rules allow unreinforced small opemngs, which are more liberal than the rules ofVIII-1. f. Fatigue Analysis The fatigue analysis of BS-5500 is very similar to that ofVIII-2. However, the SN curves are different. The fatigue exemption rules are also similar to those ofVIII-2. Only 2 SN curves are provided for a range of materials, resulting in conservatism for some materials. A major difference with SN curves of VIII-1 is that these curves have been generated from testing of specimens with ground flush joint.s. The design margins of 2 and 20 in VIII-2 are revised to 2.2 and 15. Considering all facts, the SN curves of BS-5500 are more conservative than those ofVIII-2. This British Standard also offers an alternative method based on data for welded connections (In. quiry Case 5500179). The method of CEN standard is completely different from VIII-2. The fatigue curves in this standard are based on testing of specimen with various weld details. In general, it is felt that these rules are more conservative than VIII-1 rules. Stoomwezen offers three different methods of exemption from fatigue. Other than. exemption .by satisfactory past experience, the fatigue exempt10n rules are very different from those of VIII-2. If the exemption rules are not met, a detailed fatigue analysis needs to be performed. . . The fatigue design rules ofADM are qmte different from those of VIII-2. A simplified method of exemption from fatigue is specified. Two design curves are 52 provided, one for design with low str~ss c~ncentra­ tion factor details and one for details with large stress concentration factors. This obviously is an oversimplification and results in undue conservatism for most cases. A comparison is difficult, but over a certain range the rules ofVIII-2 appear to be more conservative. A unique feature of ADM fatigue rules is that a method is offered to include the effect of creep. g. Design for External Pressure The rules of German ADM are different from those of VIII-1. Although, the, theoretical buckling pressure would be the same, the factors applied to the theoretical pressure are different. The Ger::nan code introduces a number of factors that are not m VIII-1. The rules for cone-cylinder junctions are also different. The method and the factors are unique to this code. Overall, the rules of ADM are generally more liberal. The Stoomwezen formula for allowable external pressure of cylinders is different from that ofVIII-1. A number of factors are involved that have no parallels in VIII-1. The allowable out-~f-roundness tolerances are also different. The reqmrements for stiffening ring moment of inertia are considerably less than those ofVIII-1. The formulas for spherical shells are also different. There are too many differences for a simple comparison. The rules of draft CEN standard are also different and mostly more liberal than those of VIII-1. The safety factor consists of two parts for load and ~or resistance. This results in more accurate evaluat10n and more liberal results. The concept of heavy stiffeners has been introduced in this code (a set of proposed alternative rules for VIII-1 also includes this concept). These rules are more up-to-date and mostly more liberal than those of VIII-1. The rules ofBS-5500 are similar to those ofVIII-1, but in the inelastic range they are different. The basic rules in BS-5500 are based on an out-o~­ roundness tolerance of 0.5% (this value for VIII-l 1s 1%). The resulting allowable pressures using BS5500 rules are considerably higher than those of VIII-1. The size of ring stiffeners are considerably less than those of VIII-1. The rules for spherical shells are less conservative than those ofVIII-1. 3. Design Details a. Section VIII, Division 1. The design details allowed by VIII-1 are mostly based on experience and judgment. Historically, it has been assumed that Division 1 vessels are not subject to cyclic service, unless cyclic conditions are specified and considered as required by UG-~2. !n allowing design details, the effects on cychc l.1fe generally have not been taken into account; but ~1th the design margins of the present Code and typical WRC Bulletin 435 service considerations for these vessels, there have not been any significant fatigue cracking problems. Paragraph UG-30 provides rules for attachment of stiffening rings. Such attachments are allowed by either brazing or welding. Other codes known to the authors do not allow attachment by brazing and allowing this method should be reviewed by the Code Committee. The attachment is also allowed by intermittent welding, which produces stress raisers at the start and stop points. Allowing this practice should be reviewed for vessels designed to higher allowable design stresses than presently permitted in Section VIII of ASME Code. Paragraph UG-34 provides acceptable details for unstayed fl.at heads and covers. These details are similar to those allowed by the European Codes. They include attachment by fillet welds, which are not allowed by Division 2 and some other codes. Paragraph UW-3 defines the welded joint categories. Table UW-12 contains the permissible types of welded joints. Types 1, 2, and 3 are various types of butt joints, which are also allowed by other codes. Type 4 is a double full fillet lap joint which is not allowed for the pressure boundary in some of the other codes. Type 5 is single full fillet lap joints with plug weld, which is pretty much unique to VIII-1. Type 6 is single full fillet lap joints, which are not allowed by other codes. The joint efficiencies reflect the inferiority of such details, but it is recommended that types 5 and 6 be reviewed for acceptability with lower design margins. For tapered transitions, UW-9 requires a minimum taper slope of 3:1. This is consistent with most other codes. Figure UW-13.1 provides acceptable details for attachment of heads to shells. The use of one sided fillet welds is, in general, disallowed. However, some single fillet lap welds are allowed provided adequate verlap is provided. Single fillet lap welds with pldg welds are also allowed. But weld with one plate edge offset is also allowed. These details are not provided by most other codes, but they are not prohibited either. Figure UW-13.2 provides acceptable details for attachment of pressure parts to fl.at plates to form corner joints. These details require adequate weld to develop the strength of the plate and prohibit one sided fillet welds. Figure UW-16.1 provides a number of acceptable tYPes of welded nozzles. Inside and outside pad plates are allowed. Fillet welds and partial penetration welds are allowed as well as full penetration welds. For other than full penetration attachment welds, weld strength calculations need to be made. Some of the details may result in high stress concentration factors and should be reviewed with any significant reduction in design margins. Figure UW-20 provides some acceptable types of tube-to tube sheet welds. These details are similar to those of the other Codes studied. h. Section VIII, Division 2 This division offers some acceptable design details, but certain other details may also be used, as long as they are justified by analysis or test. Paragraph AD-200 specifically allows a detail that is unique to this Code. The detail allows reduced thickness at girth seams of a cylindrical shell, under certain conditions. It has been shown by tests and analysis that the introduction of such reduced thickness girths will not appreciably affect the collapse pressure of a uniform thickness vessel. Paragraph AD-333 requires that stiffening ring attachment welds be continuous and does not allow intermittent welds which Division 1 allows. Paragraph AD-400 defines welded joint categories which are consistent with those of Division 1. However, the only tYPeS of joints allowed for main seams of the pressure boundary (categories A & B) are butt joints (seeAD-410). Joints for attachment of flanges and fl.at heads (category C) are required to be butt joints or full penetration corner joints, with some exceptions (see AD-413). For vessels constructed of materials with properties enhanced by heat treatment (AQT materials), additional restrictions apply. All welds, in general, are required to be butt welds (type 1). When the plate thickness is greater than 2", nozzle attachment welds are allowed to be full penetration corner welds (seeAD-415). or vessels in lethal service, only butt joints are allowed, except for nozzle attachment welds which may also be full penetration corner welds (see AD416). Acceptable nozzle attachment details are shown in FiguresAD-610.l,AD-612.1, andAD-613.1.All of the details call for butt welds or full penetration welds. The partial penetration attachment details of Division 1 are not allowed. However, AD-201 does allow some partial penetration nozzle attachment details (see Figure AD-621.1) for fittings which are not subject to external loadings and are not subject to local thermal stresses. It is also noted that Division 2 details all specify fairly generous corner radii. The transitions between sections of different thickness are r~uired to have a minimum 3:1 slope, unless an analysis is performed. Additional requirements are specified for nozzle neck to piping transitions (see AD-420). Details for attachment of fl.at heads are shown in figures AD-701.1, AD-701.2, and AD-701.3. Welded joints are limited to full penetration butt joints and full penetration corner joints. Partial penetration joints or fillet welded joints are not allowed. Welds attaching stiffeners or other attachments are required to be continuous (see FigureAD-912.1). Evaluation of Design Margins, Section VIII 53 Some typical attachment details are shown in Figures AD-912.1. c. Stoomwezen Some design details are specified in Sheet W0201 V, but other details are not prohibited. The acceptability of other design details are assessed from case to case by the competent body. Sheet W0301 provides acceptable weld details. Again, other weld details may be used on a case by case basis. Main pressure boundary welds may be either butt joint, or full penetration. Backing strips are not required to be removed. Joggled joints are also allowed under certain conditions. For nozzles and other attachments, fillet welds are allowed in addition to full penetration welds. For fl.at cover attachments, partial penetration welds are allowed in addition to full penetration. Additional requirements are applicable to vessels in hazardous service. Appendices 1 and 2 of Sheet W0301 contain detailed rules on weld details, which are adopted from the "International Institute of Welding" documents. There are more details provided than any other code in this study. A lot of nozzle attachment details are similar to those of VIII-1. Part 3 of Sheet W0301 contains details for studded connections and couplings. Most of details allowed by VIII-1 are included here as well as a number of other details not offered by other Codes. Part 4 includes details for flanges. Again, these are similar to those ofVIII-1, with more details provided here. Part 5 includes details for jacketed vessels. A number of details are provided which have no parallels in VIII-1. Part 6 includes tube-to tube sheet attachment details. Detailed requirements are spelled out. Part 7 includes several tube sheet to shell connection details. This code offers, by far, the greatest number of design details amongst the codes studied. To further help the designer choose the appropriate detail, Appendix 2 provides a "suitability grading" table. This table grades various details as "high," "intermediate," or "low" suitability for various service conditions, loadings, and method of examination. A great deal of information is provided in these tables making this a unique feature of this code. d. British Standard 5500 This standard offers a number of typical details. Other details may be used, as long as they meet basic design or analysis requirements. Appendix D offers a number of suggestions for details on vessels in low temperature service (below 32°F), to reduce the probability of brittle fracture. Weld neck flanges and fl.at ends are to be attached with full penetration butt welds. Appendix 4 includes recommendations for welded connections to pressure parts. Weld prep details and joint details are included which are not in ASME 54 Codes. Joggle joints are permitted with a number of qualifications. Rules for weld proximity are included which have no parallel in the ASME Codes. Several nozzle attachment details are included, mostly requiring full penetration welds. Some of the pad plate details ofVIII-1 are not included. In general a great deal more detail is specified here than there is in ASME Codes. Details of weld preps are included. For partial penetration and fillet weld attachments, very detailed requirements are specified. Details for attachment of flanges are similar to those of VIII-1, with more detailed requirements. Jacket attachment details are also included in this appendix. Details for attachment of fl.at ends are similar to those ofVIII-1, with a few additional details here. e. AD-Merkblatt This Code does not provide a lot of details. The rules are flexible enough to ascertain acceptability of various details by analysis or test. However, some recommended details are provided. AD-Merkblatt B2 offers a number of details for cone cylinder transitions. These details are similar to those allowed by theASME Codes. AD-Merkblatt B5 contains recommended details for attachment of stayed and unstayed fl.at ends. These details are again similar to those of Section VIII (except the specified values for C factors are somewhat different). AD-Merkblatt B8 includes a few details for attachment of flanges. These details do not violate any types in VIII-1 but are not as extensive. AD-Merkblatt B9, on openings, shows a few very basic types of reinforcing, including pad plates. The rules do not prohibit the use of other details. stances to meet the following requirements: • All welded joints shall be full penetration welds. • All Category A, B and C joints shall be fully radiogra phed. • All carbon and low alloy steel vessels shall be stress relieved (PWHT). • Vessels shall not be made of cast iron. Section VIII, Division 1 and Section II, Part D also includes information related to certain types of service induced degradation (Section VIII, Div. 1, Appendix HA; Section II, Part D, Appendix 6). ASME Section VIII, Division 2 vessels intended for lethal service must meet the following additional requirements: 5. Materials Requirements • All carbon and low alloy steel materials and welded joints must be impact tested. • All carbon and low alloy steel vessels must be PWHT. • Minimum permissible service temperature must be 20°F higher than the impact test temperature for each additional inch of thickness above l" (but not more than 60°F above the impact test temperature). The German Pressure Vessel Code (AD-Merkblatt) has no special requirements for aggressive or hazardous service, but the Order on Pressure Vessels, Gas Pressure Vessels and Filling Plants41 includes requirements for operation of pressure vessels and requires periodic testing by Authorized Inspector of pressure vessels belonging to Testing Group I (when used for combustible, corrosive or toxic gasses vapors or liquids), II, III, IV, VI and VII vessels.' The f. CEN Draft Pressure Vessel Standard This draft standard has very few detail sketches. A··• frequency Testing Group II, III and VI vessels is few of flange details and nozzle details are included. determined by the operator on the basis of experience wit~ the product and the method of operatio11r in the sections on design rules. These details are The testmg frequency for Testing Group IV and vt.r similar to those of VIII-1. It is not known to the vessels must be within the time periods specified in authors if design details will be added to this docuthe Pressure Vessel Order.41 ment later. . Periodic examination/overload testing of in-service pressure vessels is not addressed in Section VIII 4. Special Service Considerations of ASME Code. ASME Section VIII and The German Federal The Stoomwezen code classifies systems in three Order on Pressure Vessels include additional require~ hazard categories on the basis of the nature and the ments based on certain types of aggressive or hazard~ quantity of substances they contain. A fl.ow chart in ous service. 41 The Stoomwezen Code classifies ves- Stoomwezen document G-0701 provides a procedure sels into four hazard categories based on the nature for determining the appropriate hazard category. All and quantity of substances they contain and speci- the basic Stoomwezen rules apply to hazard category fies additional requirements for vessels in more 1 (which is thus the "Normal Case"). Hazard cathazardous service. No requirements are included in egory 2 and 3 vessels contain more hazardous subBS 5500: 1991 for special service conditions. stances. Additional requirements apply to hazard The following is a brief summary of the require~ categories 2 and 3. ments in each code related to special service consid-. The CEN/TC 54 draft Unfired Pressure Vessel erations: Standard groups pressure vessels into four Risk ASME Section VIII, Division 1 requires all unfired Categories, depending on the fluids contained (danpressure vessels which are to contain lethal sub gerous or non-dangerous), pressure and volume of WRC Bulletin 435 the vessel (the quantity of pressure times volume), ~nd type of pressure vessel. There is increasing mvolvemen~ of the Notified Body (Authorized Inspector) from Risk Category I (least risk) to Risk Category IV (most risk). AD-Merkblatt Code limits the maximum HAZ hardness in welded joints to 350 I-IV 10. The St00111 _ wezen Code limits the maximum difference between the HAZ hardness and the base metal hardness (and the weld metal and base metal hardness) to 100 HV or 150 I-Iv, depending on the base metal. (It is not clear whether this is to reduce the risk of service induced cracking or cracking during welding.) All of the codes and standards permit only those materials which have been reviewed and approved for use in pressure vessels by the appropriate code bodies or experts. The European pressure vessel codes and standards generally include national and international steel specifications. Also ASME B&PV Code has now established a Special Working Group on Non-ASTM Specifications for evaluation and acceptance of international steels. The AD Merkblatt and the Stoomwezen codes accept more structural grade plate specifications than ASME. Both AD Merkblatt and Stoomwezen codes in certain cases require testing of plates after forming. AD Merkblatt also requires identification of all weld repairs in the test certificate (similar to ASME, Division 2, but not required by SA 20 for Division 1 vessels). T~e ASME B&PV Code permits ASME SA specificat10ns for pressure parts (which generally are identi~al or similar to the ASTM specifications), materials approved by ASME Code Cases and those international materials which have bee~ approved by the appropriate ASME Code committees. The approved ASME SA specifications for ferrous materials are listed in ASME, Section II, Part A and for non-ferrous materials in Part B. Welding materials are included in Part C of Section II. The allowable design stresses and stress intensities for materials accepted by ASME Code for Divisions 1 and 2 are listed in Section II, Part D. Materials for pressure parts must conform to the specification,,, permitted by Division 1 or Division 2 as appropriate. Materials for non-pressure parts of Division 1 vessels need not conform to a material specif)_ed in Division 1, but must be of weldable quality. Materials for Division 2 vessels must meet the additional testing requirements in AM-201, test specimen heat treating requirements in AM-202 and the ultrasonic examination requirements i~ AM-203. The materials manufacturer must certify that all requirements of the materials specification Evaluation of Design Margins, Section VIII 55 and all special requirements of Division 2, Part AM to be fulfilled by the materials manufacturer have been complied with. Structural steel plates for pressure parts are limited to SA-36 and SA-283, Grades A, B, C and D. Except for flanges, bolted covers, and stiffening rings, the maximum thickness of structural grade plates shall not exceed%" (15.9 mm) for Division 1 vessels. SA-36 and SA-283 plates are not permitted for vessels in lethal service. Certification shall be in accordance with the requirements in the appropriate SA specification. Repair welding of materials by the manufacturer must be in accordance with the appropriate SA specifications. For Division 1 vessels, repair of defective materials by fabricator require acceptance by the Inspector of the repair method and the extent of repairs (UG-78). For Division 2 vessels, welded repairs of materials by the manufacturer or the vessel fabricator must meet the requirements of AF-104 or AF-750, as appropriate. BS 5500. All materials for pressure parts in BS 5500 vessels must comply with: • the British Standards listed in the Design Strength Table 2.3 of BS 5500, or • the appropriate requirements in BS 5500 and are covered by a material specification which is at least as comprehensive as the BS materials listed in the design strength tables. Several requirements (notch toughness, NDE, PWHT) in BS 5500 are related to specific groups of materials. 5 AD-Merkblatt. The materials requirements in the AD-Merkblatt code are are included in the W-series oftheAD-Merkblatt code. Most materials listed in the AD-Merkblatt code are to DIN standards, including the structural steels to DIN 17100. AD Merkblatt HP 5/241 lists the material groups in AD Merkblatt Code and the type of materials in each material group. AD-Merkblatt WO specifies additional requirements for materials and for filler metals. 41 For example, the following additional requirements are given for all product forms for Testing Group III, IV, VI and VII vessels: • The materials manufacturer shall employ equipment ensuring proper manufacture and testing of the products. • The materials manufacturer shall employ competent personnel, as well as an inspection authority for non-destructive tests, if such tests have been stipulated in the material specification. • The manufacturer shall guarantee by means of quality control and suitable records, the proper manufacture and processing of the products as 56 • • • • • • • • well as meeting all requirements of the materials specification. The suitability of the materials shall be established by an expert on the basis of the materials specification. If the suitability of the material cannot be established on the basis of the material specification, the expert shall impose additional safety conditions and appropriate tests for the product. (This is an example of where an Expert plays an important role in the AD-Merkblatt Code. For ASME code vessels generally only the ASME SA material specifications need to be complied with. Additional requirements are listed in Article M-2 for Section VIII Div. 2 vessels). Where the suitability of the material is intended for wider application, this can be done in the form of a VdTUV-Werkstoflblatt (VdTUV materials specification). The manufacturer of the materials shall prove to the expert that the requirements in 3.1.2 of AD-Merkblatt WO (i.e. the first three items in this list) have been satisfied. The works inspector may issue Test Certificate Bin accordance with DIN 50049. If the manufacturer uses, for a given product form, a material not fully refined by him, he must have in his possession certificates supplied by the manufacturer of the initial material. If DIN 50049 certificates A or C are required, tests must be carried out by an expert. In ·such cases the chemical composition of the heat, the steel making method and the method employed in the manufacture of the product must be identified. (Type A certificates are issued by a Government inspector, type B certificates are attested by a works' inspector independent from the production, and type C certificates require inspection by a third party inspector.) If the material specification does not cover re~ pair welding, it may only be done in agreement with the customer, and in presence of an expert (if an expert is required to test the product). The nature and the extent of repair welding, the type of tests to be performed on the repair welds and the test results must be indicated in a; certificate. , (Although detailed requirements for repair weld"; ing are included inASME SA-20 for steel plates, there are no requirements in SA-20 for reporting the nature and extent ofrepair welds nor the test results on repair welds). AD Merkblatt requires certification of all materials; (similar to ASME Section VIII). ·· AD-Merkblatt HP 8/1 requires one set of mechanical tests (tensile and V-notch impact tests) from eacl1 formed part having a length or diameter >4 m (13.12, WRC Bulletin 435 ft) :56 m (19.69 ft) and two sets of mechanical tests (from opposite ends) of a formed part having a length or diameter larger than 6 m (19.69 ft), with provisions for reduced amount of testing based on satisfactory prior test results. ASME Section VIII requires stress relieving of formed parts when the maximum forming strains exceed 5% (UCS-79), but there are no requirements for mechanical testing of formed parts. Stoomwezen. Stoomwezen requirements for materials are given in the M-series Stoomwezen specifications for the various types of materials and product forms. In general, only those material specifications are accepted which have been accepted in the country of issue for manufacture of pressure vessels. Approved materials are listed in the Appendices to the M-series Stoomwezen specifications. Materials are classified into various categories (A, B, B-1, C and CP) based on chemical composition and other attributes for the purpose of establishing specific requirements in Stoomwezen specifications, i.e. testing materials in their final (fabricated) condition (T 0121 and T 0221), NDE of welds (T 0110), destructive testing of welds (T 0120) and PWHT CW 0701). All carbon, C-Mn and low alloy steel plates and forgings must be supplied in the normalized, normalized and tempered or in the quenched and tempered condition. Stoomwezen Code (T 0121 and T 0221) requires mechanical testing (tensile and/or impact tests) in the following cases of materials after fabrication (forming): • Impact testing of those Category A materials which require extra impact testing, after hot forming or heat treatment (for properties), • Tensile and impact testing of Category B and B-1 materials after hot forming or heat tre7tment, • Testing for hot yield strength (when required by T0121). CEN. Materials for pressure parts shall be in accordance with the technical requirements of European Materials Standards (EN) and European Technical Approvals. Where appropriate European Standards are not available, this European Standard also permits other National or International Standards, subject to written agreement between the interested parties and approval by Notified Body (Authorized Inspector). A material may only be used for pressure parts within the temperature range for which the properties required are defined in the technical delivery specifications. Materials used for non-pressure parts welded to the vessel must be compatible with the material to which they attach and shall be supplied to a material specification covering at least the chemical composition and the tensile properties. 6. Toughness Requirements ASME Section VIII, Divisions 1 and 2. The toughness considerations in ASME Section VIII differ from those in the other international codes reviewed in this report. ASME Section VIII, Divisions 1 and 2 use an approach which is based on fracture mechanics considerations and practical considerations (based on satisfactory past experience) to preventing brittle fracture and groups materials based on their notch toughness characteristics in impact test exemption curves. ASME Section VIII, Division 1, Fig. UCS-66 includes impact test exemption curves (A, B, C and D) for determining the permissible minimum design metal temperature (MDMT) for carbon and low alloy steels based on thickness and types of materials. Curve A typically includes as-rolled, semi-killed or killed steels with relatively poor notch toughness. Curve D includes normalized and quenched and tempered fine grain steels with good notch toughness. Section VIII, Division 1 permits a 30°F decrease in the minimum metal design temperature (MDMT) if PWHT is performed on the vessel in accordance with the Code rules but not required by the Code. Fig. UCS-66.1 provides means for reducing the permissible design metal temperature (MDMT) based on a ratio of the required thiclmess to the actual thiclmess, or a ratio of the actual tensile stress in the component to the allowable tensile stress.All materials with specified minimum yield strength > 65 ksi must be impact tested. No impact testing is required on materials at metal design temperatures down to -155°F when the coincident ratio defined in Fig. UCS 66.1 is less than 0.4. (Also BS 5500, ADMerkblatt and the Draft CEN Standard have provisions for lower MDMT with reduced stresses.) The required Charpy V-notch energy values depend on thickness and the specified minimum yield strength (Fig. UG-84). The minimum average Charpy value is 15 ft-lbs for steels with yield strength Sy ::::; 38 ksi fort ::::; 2%" and 20 ft-lbs for steels with Sy ::::; 50 ksi fort::::; 1%". ASME Section VIII, Division 1, Part UCS does ffbt specify the direction of the impact test specimens in the base metal for the carbon and low alloy steels. Welded joints (WM and HAZ) must be impact tested whenever base metal requires impact testing. Recent ASME Code committee action has made Section VIII, Division 2 toughness requirements essentially the same as for Division 1 vessels. 10 Fig. AM-218.1 includes impact test exemption curves which are the same as in Fig. UCS-66 of Division 1, except that the exemption curves A, B, C and D Evaluation of Design Margins, Section VIII 57 terminate at 4 inches. The required Charpy V-notch energy values in Fig. AM-211 are the same as in Fig. UG-84.1 in Division 1. A reduction is permitted in the minimum impact test temperature based on the ration of the required governing thickness to the nominal thickness of the component for ratios down to 0.3 (Fig. AM-218.2). No impact testing is required when the design stress intensity is less than 6000 psi at temperatures above -55°F. Table ACS-1 materials in Division 2 with minimum specified UTS > 95 ksi and the high strength Table AQT-1 materials must meet 0.015" (15 mils) minimum lateral expansion for thicknesses up to 1 Vt" and 25 mils above 3" in the transverse direction for both Section VIII, Division 1 and Division 2 vessels (Fig. AM-211.2). Also certain stainless steels in both Divisions must meet the 15 mils minimum lateral expansion requirement. Certain high strength steels for Division 1 vessels (SA-508, SA-517, SA-543 and SA-592 for MDMT below -20°F and SA-645 for MDMT below -275°F) also must be drop weight tested. In Division 2 all high strength quenched and tempered steels must be drop weight tested when used at design temperatures below -20°F (except for SA-522 for all temperatures, SA-353 and SA-553 for temperatures above -325°F, and SA-645 for temperatures above -280°F). The toughness considerations in BS 5500, Stoomwezen and the Draft CEN standard are based on Charpy V-notch correlations with wide plate test results. These codes use graphs which plot reference or assessment temperatures vs. impact test temperatures for various thicknesses of materials for nonPWHT and for PWHT construction. Formulas are provided to calculate the reference or assessn1ent temperatures, which include adjustments for stress, construction categories (in case of BS 5500), PWHT of subassemblies, etc. Substantial benefits are given to stress relieved assemblies. BS 5500. Appendix D specifies toughness requirements for ferritic steels for vessels required to operate below 0°C (32°F). Figures D.3(1) and D.3(2) show graphs of impact test temperatures vs. design reference temperatures for different thicknesses for the as-welded and for the PWHT conditions. The conditions below the curves in these figures require impact testing. The application of Appendix D is limited to ferritic steels in the categories MO to M4 (carbon, C-Mn, and certain low alloy steels) in BS 5500, Table 2.3, except that rimming steels cannot be used below 0°C. Fig. D.3(2) which is intended for PWHT assemblies, permits impact testing at higher temperatures than Fig. D.3(1) (non-PWHT assemblies). The design reference temperature eR is the temperature to be used in Figures D.3(1) and D.3(2) for 58 Table23 determining the suitability of materials for the intended service. eR ::s eD + es + 8c eH 8D is the minimum design temperature (as defined in BS 5500) 8s is an adjustment depending on the calculated membrane stress. 8c is an adjustment depending on the construction category. 8tt is an adjustment where all plates incorporating subassemblies are PWHT. Table 22 lists the impact test requirements for plates, forgings, castings and tubes (from BS 5500, Table D.4.1(1)). Table22 Specified Minimum Tensile Strength Required Min. Aue. Impact Values at the Material Test Temperature <450 N/mm 2 (65250 psi) 2:450 N/mm 2 27 J (20 ft-lbs) 40 J (30 ft-lbs) AD Merkblatt. The AD Merkblatt Code requires impact testing of all carbon and low alloy steels. Impact test requirements for design temperatures down to - l0°C ( + 14°F) are generally in accordance with the applicable material specification (DIN, VdTUV materials data sheets, etc.). Impact testing generally is performed at a temperature given in the material specification which is closest to but below the lowest specified design temperature (20°C, or lower). AD-Merkblatt W 10 includes toughness requirements for low temperature service (below - l0°C, or 14°F) for pressure vessel Testing Groups III, IV, VI and VII. (No special requirements are included for vessel Testing Groups I, II and V). Table 1 in AD Merkblatt W 10 gives the lowest permissible operating temperature and maximum thickness (or diameter) for different grades of steel for Stress Categories I, II and III (100%, 75%, and 25% of maximum design stress). Table 23 is included as an example of the requirements in AD-Merkblatt W 10, Table 1, and lists the lowest permissible operating temperatures for Group 1 and Group 2 steels for Stress Categories I, II and III. The reduced toughness of the materials at the lower permissible operating temperature for stress categories II and III is intended to provide resistance to brittle fracture equal to that for Category I. Table 1 in AD-Merkblatt W 10 also includes ISO-V impact test specimen requirements, test temperature and the certification requirements. All materials in Table 1 require impact testing at the ti:>Ynn<>r~i­ tures specified in this table, or lower. WRC Bulletin 435 2 Steels in AD-Merkbl. Wl, Suitable grades ofrolled or W4, W5, WS, W12, Wl3, cast steel (as specified except killed & semiin theAD-Merkblatter killed steel below -10°C. listed in the second column) Fine grain steels to DIN Basic & elev. temp. grades 17102 & VdTUV WStE 255 through WStE 500 Werkstoftblaetter 351 and Low temp. series TStE 255 358 through TStE 420 TStE 460 TStE 500 High impact low temp. steels EStE 255 through EStE 315 EStE 355 through EStE 420 EStE 460 through EStE 500 -10 -60 -20 -70 -100°c 70mm -20°c -60 -110 -140°C -50 -100 -130°c -40 -90 -120°c 60mm 20mm 20mm -40°C -40 -40 -70 -120 -150°C 60mm -60°C -60 -110 -140 60mm -60°C -60 -110 -140 20mm -60°C Stoomwezen. The toughness rules in the Stoomwezen Code, document M 0110 apply to all parts of pressure vessels with the exception of: • Structural parts not directly welded to pressure section • Structural parts welded to the pressure section in which the primary stress is ::s50 N/mm 2 (7250 psi). The Stoomwezen Code requires two types of impact testing: • Quality testing, and • Extra testing Quality Testing. Appendix 2 to M 0110 includes a diagrams and procedures how to determine wMn the quality testing is or is not required. This testing is depends mainly on the material type and thickness. The test temperature for quality test is 20°C (68°F). Extra Testing. Diagrams 3a, 3b and 3c in Appendix 3 to M 0110, give procedures how to determine when extra testing is required for ferritic steels and at what test temperatures (e.g. at ve, vti or at another temperature). This testing depends on the type of material and operating temperature of the vessel. Figures 1 through 4 in M 0110, Appendix 3 show the relationship between the assessment temperature ve, impact test temperature Vt and the thickness de and are used to determine the impact test temperature Vt for extra testing of steels with Ni < 1.5%: • Fig. 1. As-welded condition, Rm ::s 450 N/mm 2 • Fig. 2. As-welded condition, Rm > 450 N/mm 2 • Fig. 3. Stress relieved weldments, Rm ::s 450 N/mm 2 -85°C Acc. to theAD-Merkbl. Acc. to theAD-Merkbl. listed in second column listed in second column • Fig. 4. Stress relieved weldments, Rm > 450 N/mm 2 The purpose of the extra testing is to prevent brittle fracture during the use of the vessel and during pressure testing. The risk, however, is considered acceptable when: • the material temperature is higher than the temperature for crack initiation at 0.5% plastic strain, as determined by Well's wide plate test. • the primary membrane stress is lower than the so-called Robertson plateau stress (approximately 50 N/mm2). The assessment temperature (the temperature for assessment of notch toughness) ve, is given by the following formula: Vm:min + Av1 + Av2 Vm:min = lowest design metal temperature Av1 =temperature adjustment based on stress (pressure reduction), in accordance with Stoomwezen, M 0110, Appendix 1. Av2 = 15°C when all the following under the following conditions are met: Ve= • The p~ssure part is made entirely ofC or C-Mn steel • The pressure part is spherical, or cylindrical with hemispherical heads • Shell butt welds are not stress relieved • Nozzles, nozzle reinforcements and supports welded to the pressure part have been stress relieved in accordance with W 0701 • The vessel has been subjected to a full hydrostatic test in accordance with T 0240 Av 2 = 0 in all other cases Evaluation of Design Margins, Section VIII 59 Impact testing may be required in the initial (asdelivered) condition and in final (after fabrication of the vessel) condition of the material. Impact testing of material in the initial (as-delivered) condition is required in the following cases: • Whenever the material specification requires impact testing (including test temperature). • Whenever a Quality test is required by the diagrams in Appendix 2 to M 0110. • Whenever Extra testing is required by the diagrams in Appendix 3 to M 0110. Impact tests on base metal may be transverse or longitudinal (unless specified by purchaser). No separate quality testing is required if extra impact tests are performed in the transverse direction. (See M 0110, Table 1). Tables 24a and 24b list the minimum average Charpy V-notch energy requirements for test for the quality test and for extra testing. M 0110 of the Stoomwezen Code also provides alternate procedures for material selection which parallel the assessment rules. Figure 3 (as-welded condition) or Figure 4 (PWHT condition) show the relationship between the assessment temperature Ve and thickness de for various groups of steel listed in M 0110 (B, C, D, E, F and G, depending on type of material and the impact test temperature in the material specification). The materials selected with the alternate rules from these figures will usually satisfy the assessment rules in M 0110 (including the Appendices); however, the use of these Figures is not permitted in some cases (Hazard Category 3 vessels, etc.). CEN. Appendix D in the CEN draft Unfired Pressure Vessel Standard outlines three alternative methods for establishing impact test requirements and avoidance of brittle fracture: Method 1: A code practice developed from operating experiences and applicable to all metallic materials. Method 2: A code of practice developed from fracture mechanics principles and operating experiences but applicable only to C, C-Mn and low alloy ferritic steels with specified minimum yield strength, Re :::; 460 N/mm2 (66.7 ksi). It can be applied to a wider range of thicknesses than Method 1. Method 3: Application of fracture mechanics analysis. This method is applicable to cases not covered by Methods 1 and 2 and shall only be used in agreement with the purchaser, manufacturer and the Notified Body (Authorized Inspector). It is only necessary to satisfy the requirements of any one method. Methodl: Design reference temperature TR Temperature adjustments, ts are given in Table 25. Non-welded, PWHT(l) As-welded and Percentage of Maximum Allowable Design Stress >75% :$ 100% 575% 550% 0°C 0°C +l0°C +25°C 0°c Ferritic steels, 1.5-5% Ni steels 9% Ni steels Welds for austenitic stainless steels Ferritic-austenitic stainless steels Test Ferritic steel with Ni 9% Austenitic Cr-Ni cast steel Ve< Ve 2': Ferritic cast steel Austenitic Cr-Ni steel Austenitic weld metal v < -105°C Ve 2': Other cases -105°C -105°C -105°C 31 27 43 27 43 27 39 27 Notes: 1. The values in parenthesis apply to seamless pipe. 2. Charpy V-notch test specimens shall be used; however, other test specimen are also acceptable for Quality testing at 20°C, if permitted by specifications. 60 Membrane Stress :s;50N/mm2 +50°C 0°C The required Charpy V-notch energy values Method 1 for base metal, weld metal and HAZ at impact test temperature tKV are determined from Table 26 (Table D.l, Annex D in the CEN draft Standard) using the reference thickness eB and design reference temperature tR of the component. Method 3 must be used when the wall thickness exceeds 30 mm (l.18") in the as-welded condition and 60 mm (2.36") in the PWHT condition. Method2: Parent materials, welds and HAZ shall meet the impact energy Kv in Table 27 (Table D.4 in WRC Bulletin 435 For as-welded: eB :s; 30 mm Rp 5310 N/mm2 For PWHT: eB 5 60 mm 310 N/mm2 < Rp TM 530mm 55mm None Table27 Specified Minimum Required Impact Energy, J Yield Strength of tKvfor Base Material, (10 X lOmm Non-welded tg:v for NI mm2 test pieces) & PWHT As-welded 5310 >310, 5360 >360 27 J 40J 27 J 40J 27 J Fig. D.l: PWHT condition; materials with Rp:::; 310 N/mm 2 (27 J energy requirement) and Rp > 310 N/mm 2 ( 40 J energy requirement). (This figure is the same as Fig. D.3(2) in BS 5500). Fig. D.2: As-welded condition; materials with Rp :::; 310 N/mm2 (27 J energy requirement) and Rp > 310 N/mm 2 (40 J). Fig. D.3: PWHT condition; materials with Rp > 310 N/mm 2 (27 J energy requirement). l Fig. D.4: As-welded condition; materials with Rp > 310 N/mm2 (27 J energy). Rp = 0.2% proof strength at the design temperature. 5 460 N/mm2 Limited to Tm 2': -30°C for unwelded 2" - 50°C Method 2 has no thickness limitations but has more severe toughness requirements in the non-PWHT condition than in the PWHT condition. Method3: A fracture mechanics analysis may be used to determine suitability of a particular vessel for the intended service, when agreed between the purchaser, manufacturer and the notified body, in the following cases: 1. Materials not currently covered in this CEN standard. 2. When the requirements for low temperature application cannot be adhered to. 3. Where defects outside the NDT requirements are detected. 4. Where it is proposed to use thicker material than permitted by the low temperature requirements without PWHT. D.2 D.2 D.4.a D2 D.4.b The impact test temperature tKV is determined from Figures D.1, D.2, D.3 and D.4 in the draft CEN Standard (as appropriate) using for design reference temperatures tR and reference thickness tR. Table 28 includes a comparison ofimpact test requirements for 1" (25 mm) and 1.5" (38 mm) thick as-rolled and 1.5" and 2" thick normalized SA- 516 Gr. 70 plates to be used at the maximum allowable stress at minimum metal design temperature (MDMT) of -4°F (-20°C), based on the toughness rules given in the codes and standards in this summary (Table 28). This comparison indicates more severe toughness requirements by BS 5500, Stoomwezen and the draft CEN standard than ASME, Section VIII, Division 1 for thick and for non-PWHT weldments. These codes and standards, however, do not differentiate in their toughness requirements between as-rolled and heat Table28 Material Charpy V-Notch TR TR -196°C -196°C CEN Standard) for design reference temperatures tR and reference thiclmess eB at the impact test temperature tKV. Note (1): Also applicable for equipment where all nozzles permanent welded on attachments are first welded to the nents and these subassemblies are PWHT'd before being sembled into the equipment by butt-welding, but the main are not subsequently PWHT'd. 24b Extra Testing: 27 J 27 J 40J 40J 40J None TM+ Ts TM = lowest temperature defined for each of the conditions in para. D.3.1. Ts = temperature adjustment, depending on the pressure induced principal membrane stress, as shown in Table D.2 of the CEN standard. Condition Table26 Thickness PWHT As-rolled SA 516 l" (25.4 mm) Gr. 70 1.5" (38.1 mm) No 1.5" (38.1 mm) Yes Norm. SA516 Gr. 1.5" (38.1 mm) 70 Norm. SA516 Gr. 1.5" (38.1 mm) 70 Norm. SA 516 Gr. 2"(50.8mm) 70 Yes No No Yes ASMEVIII Diu.1 15 ft-lbs & -4°F (-20°C) 15 ft-lbs & -4°F (-20°C) 15 ft-lbs & -4°F (-20°C) Testing not required Testing not required Testing not required BS5500 Stoomwezen CEN AD-Merkblatt 30 ft-lbs & -40°F 20 ft-'itl's & -22°F 20 ft-lbs & 5°F 20 ft-lbs & -4°F (-40°C) (-30°C) (-15°C) (-20°C) Requires PWHT** Requires PWHT* Requires PWHT** Requires PWHT* 20 ft-lbs & 50°F 20 ft-lbs & 39°F 20 ft-lbs & -4°F 30 ft-lbs & 57°F (10°C) (-20°C) (14°C) C4°C) Requires PWHT** Requires PWHT* Requires PWHT** Requires PWHT* 30 ft-lbs & 57°F (14°C) 30 ft-lbs & 14°F (-10°C) 20 ft-lbs & 50°F ClO°C) 20 ft-lbs & 32°F (0°C) 20 ft-lbs & 39°F C4°C) 20 ft-lbs & l4°F (-10°C) 20 ft-lbs & -4 °F (-20°C) 20 ft-lbs & -4°F (-20°C) *PWHT may be eliminated for this thickness if the material and impact testing meet the requirements in the applicable codes (see Appendix 3 for Stoomwezen W 0701, and Appendix 2 for AD Merkblatt 7/2 PWHT requirements/exemptions). **PWHT may eliminated if justified by fracture mechanics analysis. Evaluation of Design Margins, Section VIII 61 treated materials and appear to give no benefit to the improved toughness achieved by normalizing or quenching and tempering, as indicated by the lower impact test exemption curves in Fig. UCS-66 of ASME, Section VIII, Division 1. The Stoomwezen Code, however, requires the carbon, C-Mn and low alloy steels to be supplied in the normalized, normalized and tempered or in the quenched and tempered condition. 7. Fabrication Requirements a. Welded Joints Section VIII, Division 1, Fig. UW-3 defines joint categories A, B, C and D based on location in the vessel: Category A. Longitudinal welds in main shell, communicating chambers, transitions and in nozzle necks, any welded joint within a sphere, or in a formed or a flat head, circumferential welded joints connecting hemispherical heads to main shells, nozzles or to communicating chambers. Category B. Circumferential welded joints within the main shell, communicating chambers, in nozzles, or transitions in diameter, circumferential welded joints connecting formed heads (other than hemispherical) to main shells, to transitions in diameter to nozzles or to communicating chambers. Category C. Welded joints connecting flanges tube sheets or flat heads to main shell, to formed heads, to transitions in diameter, to nozzles or to communicating chambers. Category D. Welded joints connecting communicating chambers or nozzles to main shells, to spheres, to transitions in diameter, to heads or to flat sided vessels. The types of welded joints permitted in Section VIII, Division 1 are: Type 1. Butt joints as attained by double-welding or by other means which will obtain the same quality of deposited weld metal on the inside and outside weld surfaces to agree with the requirements of UW-35. Welds using metal backing strips which remain in place are excluded. Type 2. Single-welded butt joint with backing strip other than those included under Type 1. Type 3. Single-welded butt joint without use of backing strip. Type 4. Double full fillet lap joint. Type 5. Single full fillet lap joints with plug welds conforming to UW-17. Type 6. Single full fillet lap joints without plug welds. Table UW-12 in Section VIII, Division 1 lists the types of butt welded joints permitted for each joint category, limitations for each type of joint, applicable joint categories for each type of joint, and joint efficiencies based on the amount of radiographic 62 examination. Type 1 butt joints are assigned the joint efficiency of 1.0 based on 100% radiographic examination (RT), 0.85 based on spot RT and 0.70 based on no radiographic examination. The same categories of welded joints are used in ASME, Section VIII, Division 2, however, only Type 1 and Type 2 joints permitted for Division 2 vessels: Category A: Type 1 butt welded joints. Category B: Type 1 or Type 2 butt joints. Only Type 1 butt joints are permitted for the high strength quenched & tempered steels in Table AQT-1. Category C: Type 1 butt welded joints and full penetration corner joints, except that Type 2 joints are permitted in AD-416 for lethal service, and fillet welded joints are permitted in AD-711.1 for special applications. Type 1 joints are required for TableAQT-1 materials. Category D: Type 1 butt joints, full penetration corner welds, and full penetration corner welds at the nozzle neck and/or fillet welds. For Table AQT-1 materials, Type 1 joints are required for thickness :s2", but special welded joint details in Figs. AD-610.1 or AD-613.1 may be may also be used. The joint efficiency for Division 2 vessels is 1, since all butt welds must be 100% RT examined. BS 5500. BS 5500 categorizes welded joints into Types A and B based on location in the vessel: Type A. Main seam welded joints within main shells, transitions in diameter, communicating chambers, jackets and nozzles. Main seam welded joints within a flat or formed head or within a sphere. Type B. Welded joints connecting flanges, tube sheets or flat ends to main shells, to nozzles and to communicating chambers. Welded joints connecting nozzles or communicating chambers to main shells. Fillet welds or full penetration welds attaching reinforcing plates to shell and to end plates. Double welded and single welded butt joints (with and without backing strips) may be used for main seams in vessels. Permanent backing strips be used for Construction Category 1 and for Construction Category 2 (except where specifically permL1tt;ea for Construction Category 2). Single joints for main vessel welds in Construction Category 1and2 vessels are limited to 20 mm (0.79") or less, depending on the joint detail. This is restrictive than Section VIII, Division 1 which mitsASME Type 2joints (with backing for joint categories without a limitation on th:tckness, but assigns a lower joint efficiency than for joint. It is also somewhat more restrictive Section VIII, Division 2 which permits Type 2 joints (with backing strips) for Category B welds (except for the TableAQT-1 high strength materials). Generally, full penetration joint details may be used for depending on material, inspection requirements WRC Bulletin 435 intended service. This is less conservative than ASME Section VIII, Division 2 which requires full penetration corner welds at nozzle necks. Cut edges of ferritic alloy steels cut by a thermal process shall be shall be ground or machined for a distance of 1.5 mm unless approved otherwise by the Inspecting Authority. AD-Merkblatt. Requirements for welds for pressure vessels built to AD-Merkblatt Code are covered by DIN standard 8562 "Welding in the construction of vessels and containers; metal containers: welding principles." Stoomwezen. Longitudinal and circumferential welds in cylinders, cones, spheres and heads, and welds in flat heads shall be full penetration butt welds. Backing strip (if any) need not be removed from a circumferential weld, and joggled type circumferential joint may be used between two cylinders or between a cylinder and a non-conical head, providing all of the following conditions are met: • • • • • The inside of the weld is inaccessible, No crevice corrosion is expected, Static loading, Vessel is classified as Hazards Category 1, Material is not category C or CP. One sided butt welds with backing strips left in place are not permitted for ASME Section VIII, Division 2 vessels constructed of Q & Thigh strength materials. Joggled type joints in ASME Section VIII, Division 1, Fig. UW-13.l(k) are limited to%" maximum thickness for circumferential welds vessels and are not permitted for PartAQT materials in Division 2 vessels. Recommended weld details for nozzles, studded connections and couplings are given in W 0301, Appendix 1, Parts 1, 2, and 3 are essentially the same as those in BS 5500 and similar to those in ASME Section VIII, Division 1, however, a larger variety of weld details are given in the Stoomwezen Code and in BS 5500. i W 0301 Appendix 1 permits partial penetration and full penetration nozzle attachment welds. This is less restrictive than ASME Section VIII, Division 2, which requires full penetration nozzle attachment welds which must be examined by UT or RT for openings >6" dia. in shells >2112'' thick (AD-600 andAF-240). Stoomwezen requires full penetration welds and complete non-destructive examination of welds for Hazard Category 3 vessels and for all vessels made of Category C materials (which require special care in processing) and for CP materials (which are sensitive to cracking during their use). b. Impact Testing of Welded Joints ASME Section VIII, Division 1. Except as exempted in UG-20(f), the welding procedure shall include impact tests of weld metal and the HAZ when: • either base metal is required to impact tested, • the minimum design metal temperature (MDMT) is colder than -55°F and base metals are exempt from impact testing, • welds in high strength steels in Part UHT which are not subject to quenching and tempering or normalizing and tempering (except high nickel alloy filler metals in UHT-82(c)). In addition, Division 1 requires impact testing the weld metal when: • joining impact tested materials in various product forms from Table UG-84.3 or Fig. UCS-66 Curve C or D and the MDMT is colder than -20°F but not colder than -55°F unless welding consumables have been impact tested at -55°F or colder by the applicable SFA specification, • welds in UCS materials are made without filler metal and the thickness exceeds 112'' for all temperatures or when the thickness exceeds o/16" and the MDMT is colder than 50°F (UCS-67(b)), • impact tests are required for the deposited weld metal but certain high alloy steel base metals is exempted from impact tests (UHA-51). Also, the HAZ produced with or without filler metal shall be impact tested whenever the welds have any individual pass exceeding 112'' and the MDMT is colder than 70°F. Impact tests from production test plates are required for all joints for which impact tests are required for the welding procedure. Tests shall be made of the weld metal and HAZ to the extent required for the procedure test (except that no production impact tests are required for certain Part UHT materials welded with high nickel filler metals per UHT-82(c)). The impact tests from procedure qualification plates and production test plates shall meet the same requirements as base metal. ASME Section VIII, Division 2 requires welding procedure impact testing of welds and heat affected zones (HAZ) in the following cases: • When impact tests are required on the base material to be welded, • When the base material is selected from Groups IV and Vin Fig. AM-218.1, however, no impact tests are required on the HAZ when the base exempted from impact testing (AT-202). In addition, impact tests are required on all high alloy weld metal included in A nos. 6, 7, 8 and 9 of QW-422, Section IX. When the base material is required to be impact tested, Division 2 requires impact testing of welds and HAZ on production test plates for each qualified welding procedure used on each vessel. BS 5500 requires impact testing weld metal of welding procedure test plates when the operating temperature is below 0°C (32°F), but does not require impact testing ofHAZ with multi-pass welding Evaluation of Design Margins, Section VIII 63 with welding heat inputs between 1 kJ/mm (25 kJ/in.) and 5 kJ/mm (127 kJ/in.). The required minimum average impact test values are listed in Table 29 (from BS 5500, Table D.4.1(1)). Table29 Specified Minimum Tensile Strength <450 N/mm 2 (65250 psi) ~450N/mm 2 Required Minimum Average Impact Values at the Material Test Temperature 27 J (20 ft-lbs) 40 J (30 ft-lbs) The required impact test values in BS 5500 are somewhat higher than in ASME Section VIII, except for the higher strength materials/thick welds in Fig. UG-84.1. Production test plates are required when the minimum design temperature 0n is within 20°C of the design reference temperature eR and for all 9% Ni steel vessels. Impact testing of production test plates is not required in case of welds in materials less than lOmm (0.394") thick. AD Merkblatt Code requires impact testing of all carbon and low alloy steel welding procedure qualification and production test plates. Impact testing is required from each procedure qualification test plate in ferritic materials as follows (from ADMerkblatt 2/1, Table 2): For service temperatures -10°C (14°F) and above: The impact tests shall be at the same temperature as the base metal and shall meet the following minimum average ISO V-notch values: • Ferritic weld metals: the same impact test requirements as base metal. • Ferritic-austenitic, austenitic and nickel weld metals: 2:40 J (30 ft-lbs). • HAZ: 227 J (20 ft-lbs). For service temperatures below -10°C: The impact tests shall be at the lowest service temperature and shall meet the following minimum average ISO V-notch values: • Ferritic weld metals: 2:27 J (20 ft-lbs). • Ferritic-austenitic, austenitic and nickel based alloy filler metals: 2:32 J (24 ft-lbs). • The HAZ: 16 J (12 ft-lbs). The type and amount of production tests for each vessel depends on the material group, number of the cylindrical parts and the design stress level (100% or 85% of maximum allowable). The impact test requirements are (Table 1, AD-Merkblatt 5/2): Service tempeatures -10°C (14°F) and above: • Weld metal: The same as for base metal in transverse direction. (240 J ISO V-notch for ferritic-austenitic, austenitic and Ni based alloy filler metals). 64 • HAZ: 27 J ISO V-notch minimum average, at the lowest service temperature. Service temperatures below - l0°C: • Weld metal: :527 J minimum average ISO V-notch, at the lowest service temperature. (2:32 J for ferritic-austenitic austenitic and Ni based alloy filler metals). • HAZ: :516 J, ISO V-notch, at the lowest service temperature. Stoomwezen. The Stoomwezen code requires impact testing of welding procedure qualification test plates only for service temperatures below 0°C (14°F). Impact testing of production test plates is required only when extra impact testing is required. Impact tests shall be in base metal and weld metal in both the welding procedure test plates and production test plates and shall meet the requirements of M 0110. There is no requirement in the Stoomwezen code to test the HAZ. If extra impact testing is required, one production test plate must be tested for circumferential butt welds in cylinders and cones for each pressure part welded under the same welding procedure qualification and meet the impact test requirements of M 0110. Impact tests are required from base metal and weld metal. CEN. The CEN draft standard requires impact testing of the welding procedure qualification test plates and production test plates for thicknesses exceeding 12 mm (0.47''). There is no requirement for impact testing HAZ of steels in Groups 1.1 and 1.2 (low strength C and C-Mn steels). Welding procedure qualification testing is in accordance with EN 288-3 and EN 288-4. Impact testing of production test plates depends of material, thickness and design reference temperatme. Table 30 lists the impact test requirements for production test plates (from Table 12 in Part 4 of the CEN draft standard). Table30 Steel Group St 1.1, St 1.2 St 2.1, St 2.2, St 3.1, St St 4.1, St 4.2, St St 5.1, St 6, St 7.1, St7.3 Thickness, mm Impact Test Requirements e > 12 e > 12 Weld metal impacts 3 Weld metal impact tests 3 HAZ impact tests The weld metal shall meet the specified transverse impact test values for the base metal. Steels St 9.1, St 9.2 and St 9.3 shall meet 40 J minimum. The HAZ shall normally be the same as for base metal at ambient temperature, or the specified test temperature for base metal, whichever is lower. c. Inspection of Welded Joints ASME Section VIII, Division 1. The extent of NDE required in ASME Section VIII, Division 1 is WRC Bulletin 435 related to the type of service, type of material, type of weld and thickness of welded joints. Full (100%) Radiographic Inspection is required in the following cases (from UW-11): after the hydrostatic test (or after fabrication, if not accessible during the hydrotest) • All Fig. UHT-18.2 type nozzle attachment welds having inside diameter of 2" or less • All butt welds in vessel shell and heads which contain lethal substances • All butt welds in the vessels in which the least nominal thickness exceeds 1Yz" or the lesser thicknesses given in Tables UCS-57, UNF-57, UHA-22, UCL-35, or UCL-36 for the materials listed, or otherwise required by ULW-51, ULW52(d), ULW-54, or ULT-57 • All Type No. 1 butt welds in high strength heat treated steels in Part UHT • All nozzle attachment welds in high strength heat treated steels in Part UHT • All butt welds in shell and heads in unfired steam boilers with design pressures exceeding 50 psi • All butt welds in nozzles, communicating chambers, etc. attached to vessel sections or heads that are required to be fully radiographed • All Category A and D butt welds in vessel sections and heads where the design of the joint is based on the joint efficiency permitted in Table UW-12 for full radiographic examination • All butt welds joined by electrogas welding with any single pass greater than 11/z'' and all butt welds joined by electroslag welding • All welded repairs deeper than %" in materials that are required to be radiographed No specific requirements are included in Section VIII, Division 1 for MT or PT of nozzle welds in Part UCS for C, C-Mn and low alloy steels. ASME Section VIII, Division 2. The following NDE is required: Full (100%) Radiographic Examination is required for: Spot Radiography is required for all Type No. 1and2 butt welds of Table UW-12 which are not required to be fully radiographed (except for Category B and C butt welds in nozzles and communicating chambers which do not exceed NPS 10 or 1 Ya"). Ultrasonic Examination (UT) may be substituted for radiography for the final closure seam of pressurr vessel if the construction of the vessel does nd't permit interpretable radiographs. UT is required for welds made by inertia or continuous drive friction welding processes, when radiography of the welded joints is required by the Code. The ASME Code does not permit the general use of UT in place of RT as in most European pressure vessel codes. (A recent ASME Code Case 2235, however, permits the use UT in place of RT for welded joints exceeding 4" in thickness). Magnetic Particle or Liquid Penetrant Testing is required in the following cases: • When a pressure part is to be welded to a flat plate thicker than Yz" to form a comer joint the weld joint preparation shall be examined before and after welding (the remaining exposed surface) • All surfaces of weld metal buildup (UW-42) • All welds, including welds for attaching nonpressure parts to the high strength heat treated steels in Part UHT shall be MT or PT examined • All Type No. 1 butt welds, whether longitudinal or circumferential • All Type No. 2 butt welds • All Category C joints in shell thicker than 2Yz" and Category D joints greater than 611 in diameter in shell thicker than 21/z'' made by inertia or by continuous drive friction welding processes Ultrasonic Examination shall performed for: • Category C joints when the shell thickness is greater than 2Yz" • Category D joints when the opening is greater than 6" in diameter in shell greater than 21/z" in thickness • All welds made by the electroslag and electron beam welding processes (after any required heat treatments) • Corner joint constructions as shown in Fig. AD-701.3, sketches (b) and (c) Magnetic Particle or Liquid Penetrant shall be performed in the following cases: • The radially disposed surface of the opening cut in the vessel wall as shown in Fig. AD-610-1 sketches (a) and (b) • Both surfaces of welds in all full penetration corner joints • All partial penetration welds • All radially disposed surfaces cut in the vessel walls • All welds attaching non-pressure parts to pressure parts (except where reduced examination is permitted in note 1 to TableAF-241.1) Liquid Pen~rant Examination shall be performed on all austenitic chromium-nickel alloy steel welds, both butt and fillet, in vessel whose thickness exceeds%". Table AF-241.1 in ASME Section VIII, Division 2, includes a detailed summary of permitted types of welds and examination requirements for Section VIII, Division 2 vessels. BS 5500. Table 31 (Table 3.4 in BS 5500) lists three construction categories, the use of which is to be agreed between the purchaser and the manufacturer. Evaluation of Design Margins, Section VIII 65 with n11111mum yield point ;::: 370 N/mm 2 < 430 N/mm 2 ) designed up to 100% of allowable stress: Construction Category NDT 100% Spot Permitted BS 5500 Band Material None, except where limited by NDT 40 30 40 40 16 All MO&Ml M2 Austenitic SS Visual only The NDT of welded joints for final acceptance purposes depends on the Construction Category of the components and type of welded joint: Construction Category 1: The final NDE shall be performed after completion of any PWHT and shall include the following: 1. 100% RT or UT of Type A welds. 2. 100% RT or UT of Type B welds when the thinnest part welded exceeds the thickness limits in Table 32. 3. MT or PT the full length of Type Band all other attachment welds. Austenitic, MO & Ml M2 M3 M5-Ml0, inclusive 40 (1.57") 30 (1.18") 20 (0.79") 10 (0.39") Temperature Limit Lower Temp. Limit Per BS 5500 Per BS 5500 AppendixD AppendixD Per BS 5500 Per BS 5500 300°0 None None AppendixD None als group, wall thickness and the permissible design stress. RT or UT is permitted for shell butt welds. Based on satisfactory prior record, the amount of RT or UT can be reduced substantially for certain lower strength materials (which is not permitted for ASME Code vessels). A certain percentage of shell butt welds (10 or 25%) must also be subjected to MT or PT examination. Essentially all nozzle and fillet welds must be MT or PT examined. For some groups of steel the extent of the NDE can be reduced based on satisfactory prior experience. The following examples illustrate the NDE requirements in AD-Merkblatt HP 5/3 for several groups of steel listed in this document: Group 1 materials (C-Mn steels with minimum yield point 370 < N/mm 2 (53.65 ksi)): As-welded and heat treated (stress relieved) vessels designed to 100% of allowable tensile stress: • RT or UT 100% of longitudinal welds and weld intersections • RT or UT 25% of circumferential welds • MT 10% of butt welds with t > 50 mm (1.97") *See BS 5500 for listing of the various groups of steel. Construction Category 2. Construction Category 2 welded joints shall be subjected to partial NDE, except for main weld seams welded at site, which shall be 100% RT and/or UT. 1. RT and/or UT: a) of butt welds in shells, formed heads, nozzle necks, flat ends, communicating chambers and jackets b) Intersections oflongitudinal and circumferential butt welds 2. MT or PT all nozzle attachment welds and all reinforcing plate attachment welds. Construction Category 3: Only visual examination required. Both the initial assembly of components and the preparation of second side of welded joints shall be inspected and approved by the IA before welding second side. AD-Merkblatt. The type and extent of NDE required by AD-Merkblatt is listed in Tables 1, 2 and 3 of AD-Merkblatt HP 5/3 and depends on the materi- 66 Maximum Thiclmess (mm) If the results of previous NDT have revealed no serious deficiencies, UT or RT of welds in Group 1 materials can be reduced to the following: • As-welded vessels with t:::; 30 mm (1.18"): 10% oflongitudinal seams and 2% As-welded and heat treated (PWHT) vessels 1 materials designed to 85% of allowable stress: • t :::; 15 mm (0.59"): 2% of longitudinal seams, circumferential seams, and weld intersections • t > 15 :::; 30 mm (l.18"): 10% of longitudinal seams, 2% of circumferential seams, and of weld intersections Based on satisfactory prior experience, this testing can be reduced as indicated in AD-Merkblatt HP 5/3 Table 1. , As-welded (t:::; 30 mm) and heat treated. (PWHT) Group 2 steels (fine grain structural steels'y: WRC Bulletin 435 • RT or UT 100% of longitudinal seams and of weld intersection • RT or UT 25% of circumferential seams • MT butt welds 10% fort> 30 :::; 70 mm and 25% fort> 70 mm (2. 76") If the results of previous NDE have revealed no serious deficiencies, the RT or UT can be reduced for non-heat treated (non-PWHT) vessels with t :::; 15 mm and heat treated (PWHT) vessels with t :::; 30 mm to 10% oflongitudinal and circumferential seams. As-welded (t :::; 30 mm) or heat treated (PWHT) Group 3 steels (fine grain structural steels with minimum yield point ;::::430 N/mm2 ) designed up to 100% of allowable stress: • RT or UT 100% oflongitudinal seams, circumferential seams, and all weld intersections • MT butt welds 10% fort:::; 20 mm (0.79") and 25% fort > 20 mm There are no provisions for reducing the amount of RT or UT for Group 3 steels based satisfactory prior experience. Nozzle and fillet welds for all material groups require surface MT or PT examination. Additionally, UT or RT shall be performed on nozzles having I.D.;::::: 120 mm when the thiclmess of the connecting cross-sectional area exceeds 15 mm. Cracks and lack of side wall fusion are not permissible. Incomplete root fusion is not permissible for single side welded joints. Linear indications which are attributed to the separation of the materials are not permissible. The extent of NDE in the AD-Merkblatt code is somewhat more than required for ASME Section VIII, Division 1 (which has no requirement for MT or PT of nozzle welds in vessels constructed to Pijlrt 1 UCS requirements) and somewhat less than for Division 2 (which requires essentially 100% examination ofall welded joints). Stoomwezen. The materials are classified in material categories A, B, C and CP for the purpose of non-destructive examination of welds. The majority of unalloyed and low alloy ferritic steels and austenitic stainless steels belong to Category A. The extent of non-destructive examination depends on the hazard category, material category and thickness of the pressure part. Generally, more examination is required for those vessels in higher hazard categories and for thicker material. Welds in Category C materials (which require more care in processing) and Category CP materials (which are sensitive to cracking during their use) and all Hazard Category 3 vessels require 100% NDE, including temporary attachment weld locations. Butt welds in cylinders, cones, spheres, flat walls, heads and headers: Table 33 lists the extent of the examination required for material groupsAandB. The extent of limited examination must meet the following, except for tubes with outside diameter De< 325 mm (See Stoomwezen code7 for tubes with O.D. De< 325 mm): • 10% of weld length and all weld intersections • All area where an opening intersects a weld, or where the center-to-center distance between weld and opening is less than 1.5 times the diameter of that opening • One examination of each type of weld (beside the intersections listed above) • All welds in the knuckle and straight flange for dished heads The extent of extensive examination shall be the same as that for full examination (per Sheet T 0110) when: • The strength reduction factor per D 0201, D 0203 or D 0205 is higher than 0.35 • The vessel part is used in creep range • Fatigue analysis is required • Additional notch testing is required in accordance with M 0110 In all other cases, the extensive examination shall at least meet the following: • 25% of weld length • 100% of weld intersections T 0110 does not define the method of examination. It is interpreted by the authors as radiographic testing (RT) in T 0110. Table33 Wall Thickness: dd < dmin dmin $ dd $ 16 mm 16 mm < da ,,; 40 mm Full Extensive > Full Full Full Full da = wall thickness shown on construction drawings. dmin = minimum required wall thickness. Evaluation of Design Margins, Section VIII 67 Butt welds in dished and knuckled heads shall have the following NDE: 1. Wall thickness da ::; 16 mm: limited examination (as defined above) 2. Wall thickness da > 16 mm: complete (100%) examination. Materials in Category C or CP and Vessels in Hazard Category 3: The following welds require 100% non-destructive examination: • All welds in or upon pressure vessels, • All temporary field weld locations after these welds have been ground off, • Certain welds in tubes and attachment welds on tubes of Category B materials. (See Stoomwezen Code, T 0110.) Ultrasonic examination in ferritic steels: UT examination generally is performed only for thickness exceeding 8 mm (0.315"). RT examination must supplemented by UT (to the same extent as RT) in the following cases: • When 60 mm (2.36") < dr :S 90 mm (3.54") and at the same time da :S 60 mm da = actual wall thickness d = material thickness between radiation source r and recording material • Category C and CP materials, when 30 mm (1.18") < dr :S 90 mm and da :S 60 mm • If dr > 90 mm or da > 60 mm UT must be done in place or RT in following cases: • If dr > 90 mm or dd > 60 mm • Nozzle welds for Category C or CP materials when the diameter of the cutout in the vessel wall > 100 mm (3.94") and the nozzle wall thickness >20 mm (0.79") • Welds for which RT is not practical The acceptance criteria for UT examination is the same as for RT. The UT examination procedure shall meet the requirements of T 0202. Surface examination must be performed in the following cases: • All welds in or on pressure part of Category C and CP materials • All welds on the pressure part of Hazard Category 3 system • Welds for which the required RT/UT is not practical, in which case also the root pass must be examined in double butt welded joints before welding the other side. • All areas where material has been thermally removed by gouging or flame cutting (attachments, tapered edges, etc.) • Where this is otherwise specified in Stoomwezen Code, Sheet T 0110 CEN. s The NDE of welded joints depends on the 68 Risk Category (four categories), Testing Category or sub-category of welded joints (total seven categories and subcategories), group of steel, maximum thickness, and type of weld joint. It is intended that any of the testing categories will provide adequate integrity for normal purposes. The CEN draft Standard lists the following testing categories: 1, 2, 3 and 4. Testing categories 1, 2 and 3 are further subdivided sub-categories la, lb, 2a, 2b, 3a and 3b as indicated in Table 2 of the CEN standard. Testing category 4 is intended to be used only as single category for the entire vessel and shall not be combined with any other category. All butt welds in the vessel shell must be full penetration welds. RT or UT are acceptable methods for volumetric examination of the butt welds in the vessel shell. Table 34 includes a partial listing of the NDE requirements for butt welds (from Table 4 of the CEN draft Standard). All temporary attachment weld areas must be 100% MT or PT examined. All welded repairs must be 100% RT or UT examined and 100% MT or PT examined. In some respects the NDE requirements in the CEN draft standard are similar to those in ADMerkblatt code. All welds which require 100% RT or UT also require a certain amount of surface examination by MT or PT (e.g. 10%). Based on satisfactory prior experience, the amount of RT or UT for butt welds can be reduced from 100% to 10% and MT or PT for nozzle welds from 100% to 10% or 25% for certain materials. d. Tolerances ASME Code, Section VIII, Division 1 includes requirements for out-of-roundness tolerances for both internal and external pressure (UG-80). For internal pressure the differences between the maximum and minimum inside diameters are limited to 1% of the nominal diameter at the cross section under consideration. The UG-80 and Fig. UG-80.l include procedures for determining maximum plus-or-minus deviation e from the true circular form for external pressure. Most international codes have the same or similar out-of-roundness requirements. Some of the European codes have additional tolerance requirements. BS 5500 requires checking of the profile by the means of a template for vessels: • which have been subjected to fatigue analysis • vessels constructed of steel with specified minimum yield strength >400 N/mm2 (58 Ksi) Irregularity of the profile shall be checked using a template with an enclosed arc of 20° and the deviations o, referred to the mid-thickness line, shall limited to: o::; 0.05e + 0.002D (max. 25 mm), where WRC Bulletin 435 Table 34 Base Materials Longitudinal butt welds Circumf. butt welds in shell and nozzles with di> 150 mm or e> 16mm Circumf. butt welds in nozzles with di :s; 150 mm& e:s; 16mm Butt welds in spheres & heads& hemi heads to shell St ltoSt 10 RT or UT 100% MT or PT 10% RT or UT MT or PT 100% 10% RT or UT MT or PT 0 100% RT or UT MT or PT 100% 10% St 1.1, 1.2, 8.1 100% St8.2, 9.1, 9.2, 9.3, 10 100-10% St 1.1, 1.2, 8.1 100-10% St8.2, 9.1, 9.2, 10 25% St 1.1, 1.2, 8.1 10% 10%for e> 30mm 25% 10%for e>30mm 10% 10%for e>30mm 25%-10%* 10%for e>30mm 10% 10%for e>30mm 10% 10%for e>30mm 10% 100% 10%for e> 30mm 100-10%* 10% 0 100-10%* 10% 100%-10%* 10%* 100-10%* 10%for e> 30mm 10% 10% St 1.1, 8.1 0 0 0 0 10% 0 10% 0 0 25% 10% 10% 10%for e>30mm 0 *The second number indicates the extent of examination permitted after successful experience (as defined in CENtrC 54 draft standard for Unfired Pressure Vessels, Part 5). ois the maximum local irregularity; e is the plate thickness, mm; Dis the outside diameter of the shell, mm. BS 5500 also limits the straightness of the vessel to 0.3% of either the total cylinder length or of any individual 5 m (16.4 ft) length of the vessel. AD-Merkblatt specifies limits for out-of-roundness and local under thickness deviations (ADMerkblatt HP 1) but does not include limitations for peaking or local flat areas. The past experience by the authors is that the AI (TUV) will require propf (by calculations or strain gage measurements) during overload testing that all stresses will remain in the elastic range after the first overload test. The Stoomwezen Code (T 0230) gives procedures for measuring out-of-roundness of cylindrical shells and for determining local deviations (for vessels subjected to external pressure), but does not specify maximum permissible deviations. The CEN draft standard includes tolerance limits for middle line alignment, surface alignment, out-of-roundness for vessels subject to internal pressure, peaking at welded joints, local area underthickness, and dished heads. The maximum offset limits for middle line alignment and surface alignment depend on type of shell seam (longitudinal or circumferential), location (cylinder, cone, dished head, sphere) and thickness. The straightness of the vessel shall not exceed 0.5% max. of total cylindrical length (vs. 0.3% in BS 5500). Peaking shall be checked by a 20° gage length and shall not exceed 2% of the length of the gage. This can be increased by 25% ifthe length of the peaking does not exceed :Y,. of the length of the shell between two circumferential joints, with a maximum of 1 m. Limitations are also specified for peaking associated with flat areas adjacent to the welded joints. e. Postweld Heat Treatment ASME Section VIII, Division 1 requires PWHT of vessels in the following cases: • All carbon and low alloy steel vessels in lethal service. • Cold formed carbon and low alloy steel plates when the extreme fiber elongation is more than 5% and any one of the following conditions exist: -the vessel will contain lethal substances -the material requires impact testing -the thickness of the part before cold forming exceeds%" -the material during forming is in the range 250°i,and 900°F • When the nominal thickness (including corrosion allowance) exceeds the the thicknesses given in Tables UCS-56, UHT-56, UHA-32 and UNF-56. PWHT shall be at the minimum temperatures/temperature ranges given in these tables for the materials listed • When the MDMT is colder than -55°F, and for vessels installed at a stationary location, the coincident ratio defined in Fig. UCS-66.l is 0.4 or greater • When welding ferritic materials greater than Ya" with the electron beam process Evaluation of Design Margins, Section VIII 69 • When welding P-No. 3, P-No. 4. P-No. 5 and P-No. 10 materials of any thickness using the inertia or continuous drive friction welding process Electroslag welds in ferritic materials over l Yz" thick and eletrogas welds in ferritic materials with any single pass greater than 1Yz" shall be given austenitizing heat treatment. ASME, Section VIII, Division 2 requires PWHT all ferrous materials in welded pressure vessels or parts when: • the nominal thiclmess, including corrosion allowance, of any welded joint exceeds the limits in TablesAF-402.1andAF631.1 • the vessel is designed for lethal service • when welding ferritic materials greater than Ys" thick with electron beam welding • when welding P-No. 3, P-no. 4, P-No. 5A, 5B and 5C, P-no. 6, P-No. 7 (except Type 405 and Type 410S) and P-No. 10 materials using the inertia and continuous drive friction welding process. BS 5500. The PWHT requirements in BS 5500 are similar to those in ASME Section VIII. BS 5500, however, permits as-welded vessels of C and C-Mn steels up to 35 mm (1.38"). This can be increased up to 40 mm (1.57") provided the materials are impact tested to meet 27 J (20 ft-lbs) at -20°C (-4°F). PWHT of welded assemblies is required in the following cases: • Ferritic steel vessels designed to operate above 0°C (32°F), when the thickness of welded joints exceed the limitations in BS 5500, Table 4.4.3.1. When agreed to by the purchaser, manufacturer and the Inspecting Authority, the thickness limits in Table 4.4.3.1 may be exceeded, providing a fracture analysis is performed in accordance with Appendix U and PD 6493 • Ferritic steel vessels designed to operate below 0°C (32°F) when PWHT is necessary in accordance with Appendix D • Vessels intended to operate in service which could cause stress corrosion cracking • Where specified by the purchaser AD Merkblatt. PWHT depends on the materials group and thickness. Normally, C and C-Mn steel welded assemblies with thickness exceeding 30 mm (1.18") must be PWHT. AD-Merkblatt HP 7/2 permits a maximum thickness of 50 mm (1.97") without PWHT for material groups 1 and 5.1 providing the following additional requirements are met: • Impact tested fine grain structural steels to DIN 17102, 17178, 17179, etc., to meet 31 J, transverse, at 0°C (32°F). • Vessels of simple geometric shape (sphere, cylinder) • 100%NDE • Test stress during overload test 2::0.85 Remin 70 (specified minimum yield strength) at room temperature • Special brittle fracture investigation is performed • Components with nozzles and welded on parts are stress relieved Stoomwezen Code, W 0701 requires PWHT of welded joints in carbon and low alloy steels when the thiclmess exceeds 32 mm (1.26"). The PWHT temperatures and times generally are in accordance with published material specifications. If no information is given on PWHT, the following requirements shall be met: • Minimum PWHT temperature for unalloyed and low alloy steels: 600°C (1112°F) • Permissible temperature variation: +/- 20°c (±36°F) • Minimum hold time: 120 de, sec., with a minimum of 1800 sec. (0.5 hr) • Maximum hold time: 7200 sec. (2 hrs) The following examples list PWHT requirements for Material Groups 1, 2, 3 and 5.1 (from AD-Merkblatt are from HP 7/2). PWHTcan be omitted when all the following conditions have been met: Materials Group 1 (C and C-Mn steels with specified minimum yield point <370 N/mm2): s < 30 mm (1.18"): No additional requirements. (s =thickness at welded joint) s > 30 :;::; 50 mm (2"): Fine grain structural steels to DIN 17102, 17178 and 17179, and other steel grades which meet the same minimum impact test requirements (as listed in DIN 17102, etc.). s > 38 :;::; 50 mm (2"): All steels with specified impact test requirement 2::31 J (23 ft-lbs), transverse, at 0°C (32°F). Simple geometrical shape (sphere, cylinder); 100% NDE; max. overload test stress 2:: 0.85 UTS; brittle fracture assessment; nozzle assemblies shall be PWHT. Materials Group 2 (fine grain structural steels with specified minimum yield point 2::370 <430 N/mm2): s :;::; 30 mm: Fine grain structural steels to DIN 17102, 17178 and 17179 and other steel grades which meet the same impact test requirements. Materials Group 3 (fine grain structural steels with minimum yield point 2:: 430 N/mm2 ): s :;::; 30 mm: All grades in this category. Materials Group 5.1 (fine grain structural steels to DIN 17102, 17178 and 17179 with specified minimum yield point 2::370 N/mm 2): s :;::; 30 mm: All grades. s > 30 :;::; 38 mm: Fine grain steels to DIN 17102, 17178 and 17179 and all grades which meet the same min. impact test requirements. s 2:: 38 :;::; 50 mm: All grades with min. impact requirement 2:: 31 J, transverse, at 0°C. Simple geometrical shape (sphere, cylinder); 100% NDE; overload test stress 2:: 0.85 UTS; brittle fracture assessment; nozzle assemblies PWHT. Group 5.4 materials (3.5, 5 and 9% Ni steels) need not be PWHT with thiclmess s :;::; 50 mm, providing they are welded with austenitic or Ni based alloy filler metals. (The maximum thickness without PWHT for these steels is 30 mm if welded with same type filler metals as the base metal). The PWHT temperatures and hold times in ADMerkblatt Code generally are in accordance with the material specifications. The PWHT temperatures for C and C-Mn steels generally are lower than those in ASMECode. WRC Bulletin 435 de is the effective thickness of weld, mm (as defined in W 0701 for evaluating PWHT requirements). PWHT is required when the wall thickness de > 32 mm (1.26"). However, PWHT may be required also for lesser thicknesses, e.g.: • Steels with carbon equivalent Ceq > 0.45% or carbon content C > 0.25% • Fine grain steel with C > 0.23% • Fine grain steels with Ceq > 0.55% or specified minimum yield strength Re > 450 N/mm2 and hardness difference between HAZ and base metal or HAZ and weld metal ~H > 150 HV Welded joints with de < 40 mm in fine grain steels need not be PWHT when all the following requirements are met: • • • • c:;::; 023% Ceq:;::; 0.55% Re:;::; 370 N/mm2 Base metal meets Charpy V-notch impact test energy 31 J at 0°C or 27 J at -50°C and weld metal 2:: 31 J at 0°C and the vessel is subjected to hydrostatic test. The CEN drafi Standard, Part 4 (Table 18) requires PWHTofwelds in low strength C and C-Mii steels (St 1.1 and St 1.2), normalized fine grain C-Mn steels (St 2.1), C-Mo steels 16Mo3 (St 1.2), MnNi and Ni bearing low alloy steels (St 9.1 and St 9.2, except St 9.3, 9% Ni steel X8Ni9), when the thickness exceeds 35 mm (1.38"). The PWHT temperature for thicknesses over 35 mm is 550-600°C (1022-ll12°F) for St 1.1, St 1.2 and St 2.1, 550-620°C (1022-1148°F) for St 1.2, and 530-580°C (986-1076°F) for St 9.1 and St 9.2, with the hold times listed in Table 35. ::;;35 >35 :s; 90 en, nominal thickness of weld, is defined in the CEN draft standard, Part 4. The 35 mm maximum thickness without PWHT may be exceeded when justified by fracture mechanics analysis. The PWHT temperatures and hold times in the Draft CEN Standard are lower than those required by ASME Code. If repair welding is performed after PWHT or after the hydrostatic test, these operations shall be repeated. f. Third Party Inspection The Authorized Inspector (AI) has the responsibility for monitoring the ASME Section Vlll Division 1 vessel Manufacturer's Quality Cont;ol System as well as for the duties in UG-90(c). The AI for ASME Section VIII, Division 2 vessels has similar responsibilities. These are outlined in AG-303 and AI-102. Some of the Al's duties include verification that: • the manufacturer has a valid Certificate of Authorization and is working to a Quality Control System • the applicable Design Report, User's Design Specification, drawings and other related documents are available • the materials comply with the requirements of the Code • all welders, welding operators and welding procedures have been qualified • all heat treatments, including PWHT, have been performed • material imperfections have been properly repaired and examined • the various inspections and tests have been performed and that the results are acceptable • the required marking is provided and that the name plate is properly attached • the manufacturer has maintained proper records The inspector has to perform internal and external inspections and witness the pressure tests and sign the Certificate of Inspection on the Manufacturer's Data Report, but he does not need to verify the design nor perform or supervise NDE or other types of tests (as stipulated in some European pressure vessel codes and standards). BS 5500. The Inspecting Authority (IA) is responsible for ve:Nit:ying that: • all parts have been designed in accordance with requirements of BS 5500 an as specified by the purchaser • that the vessel has been constructed and tested in accordance with BS 5500 and the purchaser's options (as listed in Table 1.5 of BS 5500) The specific duties for IA during manufacturing and testing are given in Table 5.1 of BS 5500 and are summarized below: • Correlation of material certificates with materi- Evaluation of Design Margins, Section VIII 71 • • • • • • • • • • als and check for conformity with material specifications Identification of material and witnessing of transfer of identification marks at manufacturer's facility Examination of cut edges and heat affected zones (Category 3 components). Approval of welding procedures Approval of welders and operators Examination of seams for welding, including dimensional check, weld preparations, tack welds, etc. (Category 3 components) Inspection of second side of weld preparation after preparation of first side and root cleaning (Category 3 components) Examination ofNDE reports and check compliance with the agreed procedure and acceptability of any weld defects. Examine heat treatment records and check compliance with the agreed procedure Witness pressure test and, where necessary, record the amount of permanent set Examine completed vessel before shipping. Check marking. The German Pressure Vessel Code 6, 41 specifies that certain Testing Group vessels may not be put in service until the Authorized Inspector has subjected the vessel to an initial test (inspection) and an acceptance test (inspection) and has certified that he has found it in a serviceable condition. These vessels belong to Testing Groups III, IV, VI and VII. The initial test consists of a preliminary test, a structural test and a pressure test. The acceptance test consists of a service test, an equipment test and an installation test. (See Appendix 2.) The requirements for involvement of Inspection Authority (or Authorized Inspector appear throughout the AD-Merkblatt Code. Some of the requirements are: • NDE of materials, if such tests are stipulated in the material specification (AD-Merk.blatt WO) • Verification of materials, testing, marking and certification of materials for vessels belonging to Testing Groups III, IV, VI and VII (W-series AD-Merkblaetter) • Transfer of marking of materials subject to construction inspection by the AI • Review and approval for all modifications and repair work on finished components and vessels after NDE and after the construction inspection or partial construction inspection (AD-Merkblatt HP O) • Verification of welding procedure qualification tests for vessels subject to construction inspection (AD-Merk.blatt HP 2/1) • Witnessing tests on production test plates for vessels subject to construction inspection (ADMerkblatt HP 5/2). • Witnessing and supervision of NDE of vessels 72 subject to construction inspection (AD-Merkblatt HP 5/3) • Review of heat treat procedures for parts and vessels subject to construction inspection (ADMerkblatt HP 7/1, HP 7/2) • Testing by Authorized Inspectors; Initial Test (TRB 511), Pressure Test (TRB512), Acceptance Test (TRB 513), Repeat Tests (TRB 514) Sto01nwezen. The requirements for involvement by the Notified Body (Inspection Authority) appears throughout the Stoomwezen Code. The following lists some of the involvement by Notified Body during manufacturing: • Approval of material specifications and certification of materials (M 0201throughM0803) • Supervision during fabrication, examination of Hazard Category 2 and 3 vessels (T 0101) • NDE of welded joints (T 0110, T0112). • Destructive testing of welds CT 0120) • Mechanical testing of materials CT 0121) • Welding procedure qualification (T0210 & T 0211) • Welders' qualification CT 0215) • Testing of materials in final condition CT 0221) • Out-of-roundness measuremens CT 0230) • Pressure test CT 0240) • Certification of vessels CT 0270) Drafi CEN Standard. The Notified Body (Al) is responsible for verifying (where applicable): l. that all parts of the vessel have been designed in accordance with the requirements of the Draft CEN standard to the designated design criteria. 2. that the vessel has been constructed and tested in accordance with this standard. The extent and level of participation by the Notified Body depends on: • Risk categories (I, II, III and IV) • The testing category (1, 2a, 2b, 3a, 3b, and 4) • The module of conformity assessment (A through Hl) The vessel manufacturer has an option of selecting between a module of conformity assessment involving a certified quality system and one which does not; however, it shall be agreed with the purchaser and communicated to the Notified Body. In the case of conformity assessment module involving a certified quality system, there is a reduced involvement by the Notified Body. Table 36 (Table F.2-1 in Section 5 of the CEN draft Standard) lists the applicable conformity assessment modules for the various hazards categories without and with a certified quality system (QS). The various conformity assesment modules are described in the European Pressure Equipment rective (PED). 42 The Notified Body is responsible for WRC Bulletin 435 I II III IV A Al Bl+F A Al B +Cl DlorEl Bl+ D,orH DlorEl B + E QS = quality system. the following activities when the conformity assessment module Hl has been selected: • Assessing the vessel manufacturer's quality system • Evaluating any changes to the approved system • Performance of surveillance and audits Table H-1 in Section 5 of the CEN draft Standard provides a detailed summary of the activities and participation by manufacturer and Notified Body for individual pressure vessels with respect to conformity assessment modules A, Al, Dl, El, Bl+ F, H, Bl + D, G and Hl. Table H-2 lists the involvement by manufacturer and Notified Body for serially produced pressure vessels with respect to the various conformity assessment modules. The Notified Body is responsible for inspection operations related to: • Design, purchase and subcontracting specifications • Material inspection • Welding and forming • Production tests • PWHT • NDE • Pressure test • Safety devices • Conformity assessment least 30°F above the minimum MDMT to minimize the risk of brittle fracture. The hydrostatic test pressure for ASME Section VIII, Division 2 vessels is at least 1.25 times the maximum allowable working pressure in every part of the vessel. The pneumatic test pressure shall be at least 1.15 times the maximum allowable working pressure (Articles T-3 and T-4). The primary general membrane stress intensity Pm shall not exceed the 90% of the specified minimum yield strength Sy at the test temperature for the hydrostatic test and 0.8Sy for the pneumatic and hydro-pneumatic test. The test stresses shall also satisfy the requirements: Pm+ Pb 5 1.35 Sy for Pm 5 0.67Sy and Pm+ Pb:$ 2.15Sy - 1.2 Pm for 0.67Sy <Pm:$ 0.9 Sy Pb = Primary bending stress intensity Both Division 1 and Division 2 include the following considerations for the metal temperature during overload testing to minimize the risk of brittle fracture: • The Code recommends that the minimum metal temperature during the hydrostatic test be maintained at least 30°F above the minimum metal design temperature (MDMT), but need not exceed 120°F. • The Code requires that the metal temperature during the pneumatic test be maintained at least 30°F above the minimum MDMT), but need not exceed 120°F. BS 5500. The test pressure for hydrostatic, pneumatic and combined hydrostatic/pneumatic tests shall not be less than the following: Category 1 and 2 construction: Pt = 1.25 p (fjft) (t/(t c)] Category 3 construction: 8. Overpressure Testing Requirements The hydrostatic test pressure for ASME Section VIII, Division 1 vessels is at least 1.5 times the maximum allowable working pressure in every part of the vessel multiplied by the lowest ratio of the allowable design stress at the test temperature to the allowable design stress at the design temperature. (Recent changes to the Division 1 requirements have reduced the hydrostatic test factor from 1.5 to 1.3). The Code does not specify the maximum upper limit; however, vessels with visible permanent distortion may be rejected by the Authorized Inspector (Al). The pneumatic test pressure shall be at least 1.25 times the maximum allowable working pressure multiplied by the lowest ratio of the allowable design stress at the test temperature to the allowable design stress at the design temperature. The metal temperature during the pneumatic test shall be at Pt = 1.25 P (fafft) (t/(t - c)], or Pt = 1.5 p, whichever is higher p = design pressure fa = nominal design strength (stress) at test temperature ft = nominal design strength (stress) at design temperature t = nominl!! thickness of the section under consideration c = corrosion allowance The general membrane stress in any part of the vessel during test shall not exceed 90% of the minimum specified yield strength of the material. AD Merkblatt. The maximum overload test pressure at the highest point in the vessel shall be: Hydrostatic test: 1.3 X p for the following materials: steel and malleable non-ferrous metallic materials, cast steel, cast iron. Evaluation of Design Margins, Section VIII 73 Two alternates are given for pneumatic test: c = corrosion allowance 1. The same test pressure as for hydrostatic test (and the applicable materials) when the pressure vessel is located in a place where nobody is in the immediate danger while the vessel is subjected to the test pressure (i.e. explosion proof test chamber, water basin with the appropriate precautions to prevent flying objects). 2. Pp= l.lp. The temperature of the shall meet all the following requirements: Stoomwezen. The required hydrostatic test pressure Pt is given in the table below (Table 37): The maximum allowable hydrostatic test pressure Pt is: • Pt = 13.5 Pdi for dished parts with pressure on the convex side the lowest "adjusted design pressure" at P dl test temperature for each pressure part, calculated without deducting the corrosion allowance • Pt = 14 Pdi for all other cases The minimum test temperature must be at least 5°C (9°F) above the freezing point, but no lower than l0°C (50°F). Pneumatic test (including pneumatic/hydrostatic test). The required and allowable test pressures for are: • The required test pressure Pt = 10 Pd • The allowable test pressure Pt = 10 Pd1 Pdi = the lowest "adjusted design pressure" at 20°C (68°F) for each pressure part, calculated without deducting the corrosion allowance. Toughness evaluation of the vessel shall be made in accordance with Stoomwezen M 0110 if the test temperature reduces to below 15°C (59°F) during the pneumatic test. CEN. For Testing Category 1, 2 and 3 vessels, the hydrostatic test pressure Pt shall be the higher of: Pt = 1.25 Pd (falft), or Pt= 1.43 Pd Pd = the design pressure of the vessel part fa = nominal design stress at test temperature ft = nominal design stress at design temperature For Testing Category 4 vessels the test pressure shall be the higher of the following: Pt = 1. 75 Pd (fafft)(e/e-c) for St 1.1 steels Pt= 1.6 Pd (fafft) for St 8.1 steels e = nominal thiclmess of the section under consideration 1. It shall be above the freezing point 2. It shall be 10 K (18°F) below the atmospheric boiling point 3. It shall be of sufficient temperature to avoid brittle fracture The pneumatic test pressure shall be: Pt = 1.1 Pd (fa/ft) Special consideration must be given to vessels subjected to pneumatic test, e.g.: • MT or PT of all welds which have not been subjected to 100 RT or equivalent inspection as required for the main seams. • Test temperature at least 25°C above the impact test temperature required by this CEN standard for vessels which have not been previously subjected to a hydrostatic test at a pressure exceeding the pneumatic test pressure. • Absolute necessity of avoiding brittle fracture. • Remote monitoring during the test. of the quality system during the initial approval phase of the facility and review of the quality system during the production phase. The CEN draft Standard lists various modules for conformity assessment. The vessel manufacturer has the option of selecting an appropriate module of conformity assessment; however, it shall be agreed with the purchaser and communicated to the with the Notified Body. In the case of conformity assessment module involving a certified quality system, there is a reduced involvement by the Notified Body. The vessel manufacturer must have a quality management system for vessels fabricated in accordance with the conformity assessment modules listed in Table 38. Table38 Module of Conformity Assessment Quality Management System E D H ISO 9003 (EN 29003) ISO 9002 (EN 29002) ISO 9001(EN29001) 10. Summary and Conclusions 9. Quality Control System ASME Section VIII, Divisions 1 and 2. Both, Division 1 and Division 2 require that the vessel manufacturer or assembler have and maintain a Quality Control System which establishes that all Code requirements, including material, design, fabrication, examination (by the manufacturer or assembler) and inspection (by the Inspector) will be met. The requirements for Quality Control System for manufacturers of Division 1 vessels are included in Section VIII, Division 1, Appendix 10. The requirements for Quality Control System for manufacturers of Division 2 vessels are included in Section VIII, Division 2, Appendix 18. No specific requirements are included in BS 5500 for quality control. Although quality control is mentioned in the ADMerkblatt Code, there are no specific requirements for a quality control system in the Code. CEN. The materials manufacturer shall maintain a quality system in accordance with EN ISO 9002 and obtain certification for this system from a Notified Body (a third party authorized to perform the conformity assessment tasks). Materials manufacturers are subject to inspection This part of the Phase 2 included a review of four European pressure vessel codes and standards and a comparison withASME Section VIII, Divisions 1 and 2. The following are the observations and conclusions from this study: Design Margins All the European codes and standards included in this study have lower design margins at temperatures below the creep range than in both Section VIII Division 1 and Division 2. For un-alloyed and low alloy steels, all these codes and standards use a factor of 1.5 on the specified minimum yield strengt)l at design temperature and different factors on tB:e specified minimum tensile strength. The factors on tensile strength are: • 4 at design temperature for ASME, Section VIII, Division 1 • 3 at design temperature for ASME, Section VIII, Division 2 • 2.35 at room temperature for BS 5500 for Construction Categories 1 and 2 (5 for Construction Category C and C-Mn vessels) • 2.27 at room temperature for Stoomwezen (with elongationA5 > 10%, otherwise the factor is 4) • 2.4 at room temperature for the Draft CEN Standard (1.875 for the Alternative Design Basis, which was included in the July 1995 CEN/TC 54 Draft Standard for Unfired Pressure Vessels but not in the May 1998 Draft.) The AD-Merkblatt Code specifies factors on ultimate tensile strength only for gray cast iron and for copper and its alloys. Pd 14 WRC Bulletin 435 The allowable design stresses for stainless steels and non-ferrous alloys are the same as for C, C-Mn and low alloy steels, or somewhat different, but in any case, more liberal than those for both ASME Section VIII, Divisions 1and2. No actual comparisons were made on allowable design stresses in the creep range; however, the allowable stress basis for each code and standard is included in the report. All the European codes and standards addressed in this report (except BS 5500) require adding the material under-thiclmess tolerance to the calculated shell thickness. Design Rules The basic formulas for the design of shells are almost the same in various codes and standards. These formulas, being based on simple shell theory, cannot be modified or improved much. However, the design rules for special components and for areas of discontinuity vary a great deal from one code to another. In most European Codes the rules for design of cone-cylinder junctions are based on limit analysis or other analytical methods. Such rules in the ASME Codes are scattered in various parts and could be improved by the use of available analytical methods. However, the existing rules are conservative and have resulted in no known problems. The existing rules in the ASME Codes for the design of dished heads need to be improved. Alternative rules being considered by the ASME Code Committee will significantly improve these rules. Although the present rules have a rather arbitrary basis, resulting in a varying safety factor over the range of geometric parameters, they are quite conservative. Over the most of the range of geometries, the calculated thiclmess of ASME Code (Div. 1 and Div. 2) heads are greater than those of European Codes investigated in this report. Failure by buckling should be included in the ASME rules and the range of geometries extended to thinner shells. The formulas for design of fl.at covers are similar in all codes. However, the value of "c" factors vary significantly. Analytical tools are available to arrive at more accurate values of the "c" factors. Various code organizations should cooperate to specify consistent values. Flat covers by ASME rules, would be generally tmcker than those designed by European Codes, mostly due to lower allowable stresses. The rules for design of bolted flanges in theASME Codes are conservative. A revised set of rules is being developed which has not been studied for this report. It is difficult to assert that the ASME rules are consistently conservative relative to European Codes and Standards. Opening reinforcement rules vary a great deal also. The ASME pressure area method has been extensively used with excellent results. However, these rules may be overly conservative. The Euro- Evaluation of Design Margins, Section VIII 75 pean Codes and Standards have recently used more analytical approaches, such as limit analysis. A review of ASME Code approach to penetration design is desirable. There are no rules for fatigue exemption or fatigue analysis in ASME Section VIII, Division 1. It would be desirable to provide some guidelines. The fatigue curves in Section VIII, Division 2 are based on smooth specimen tests, and the fatigue analysis rules require evaluation of stress concentration factors at discontinuities. The Section VIII, Division 2 rules could possibly be unconservative in some cases. A thorough review and update of these rules is needed. The European rules are based on fatigue tests on weldments and generally are more up-todate and comprehensive. The rules for external pressure in the ASME Codes are conservative. A set of alternative rules, based on latest test data, are included in Code Case 2286. In general, the present ASME buckling rules are more conservative than those of the European Codes. Overall, the comparisons of this report indicate that design rules of the ASME Code Section VIII (specially Division 1) are considerably more conservative than those of the European Codes studied. However, some ASME design rules are out-of-date, do not have adequate analytical basis, and provide a non-uniform margin of safety from one component to another over the range of geometries. The ideal situation, of course, would be for all vessels to have uniform safety margins for all components. Although this will not be totally achievable, a more uniform safety margin should be strived for. Special Service Considerations Both ASME Section VIII, Division 1 and 2 require vessels in lethal service to be fully radiographed (Category A, Band C joints for Section 1 vessels, and all butt joints in Division 2 vessels, as specified in Division 2). AD-Merkblatt Code does not specifically address lethal or hazardous service, but requires periodic testing (NDE and pressure testing) of Testing Group I vessels (when used for combustible, corrosive or toxic gases, vapors or liquids) and all Testing Group II, III, IV, VI and VII vessels. 41 Stoomwezen Code classifies all vessels in four hazards categories, with the most severe (Category 3) requiring 100% NDE. The Draft CEN Code also classifies its vessels in four risk categories, with the highest risk category vessels requiring additional quality control and the most involvement by the Notified Body (Inspection Agency). Material Requirements All codes and standards only permit materials which have been approved by the appropriate code bodies or experts. The AD Merkblatt and Stoowezen Codes accept more and higher strength structural grade plates than ASME, but both these codes have some additional requirements beyond those in Sec76 tion VIII, Divisions 1 and 2. Some of the requirements are: AD-Merkblatt. • Materials used for pressure vessels belonging to Testing Groups III, IV, VI and VII are subject to additional special requirements (as specified in AD-Merkblatt W)) • The suitability of filler metals and consumables for pressure vessels belonging to Testing Groups III, IV, VI and VII must be established by an expert's report • Certification is required of all materials (similar to Section VIII, Division 2) • Use of Inspection Authority for non-destructive tests (if such tests are required in the material specification) • The type and extent of repair welding and the type of tests performed on repair welds must be indicated in the test certificate • Generally, all materials are subject to notch toughness requirements at the lowest specified design temperature, however, materials used at service temperatures above -10°C (14 °F) generally are impact tested at room temperature • Additional testing of plates after forming. One set of mechanical tests (tensiles and impact tests) from one end of each formed part with length or diameter greater than 4 m (13.12 ft) and both ends from of a formed part having a length or diameter greater than 6 m (19.69 ft) Stoomwezen: • Only materials specifications reviewed and approved by the Competent Body can be used for pressure vessels. Materials must meet any additional requirements established by the Competent Body • All non-alloyed and low alloy steel materials to be supplied in the normalized, normalized and tempered, or in the quenched and tempered condition • Ferritic materials are subjected to impact testing at 20°C for quality requirements • Stoomwezen Categories B and B-1 materials must be impact tested after hot forming and all heat treatments (including PWHT) CEN: • European technical approval of material manufacturers • Review and approval of the material manufacturer's quality system by the Notified Body (Authorized Inspection Agency) ASME Section VIII covers the requirements pertaining to vessels constructed of specific materials groups in separate parts of the Code (e.g. Parts UCS, UDA, UHT, etc. Division 1 and Part AQT, etc. in Division 2), whereas the European codes and standards assign materials into various materials groups and WRC Bulletin 435 cover the requirements pertaining to specific materials groups throughout the codes and standards. Toughness Requirements The toughness requirements in the European pressure vessels codes and standards differ from those in ASME Section VIII. The toughness requirements in BS 5500, Stoomwezen and the CEN draft pressure vessel standard generally are based on Charpy Vnotch correlations with wide plate test results. These codes use graphs of vessel reference or assessment temperatures vs. impact test temperatures for different thiclmess materials for non-PWHT and for PWHT construction. The impact test temperature depends on the calculated reference or assessment temperature, thic~ness and whether the welded assembly is stress reheved. Formulas are provided to calculate the reference or assessment temperatures which include adjustments based on calculated membrane stress, construction categories (in case of BS 5500), PWHT of vessel assemblies, etc. Substantial benefits are given to PWHT. B~ .5500 includes toughness requirements only for ferntic steels required to operate below 0°C (32°F). ?'he required minimum average impact values at the impact test temperature are 27 J (20 ft-lbs) for steels with specified minimum UTS < 450 N/mm2 (65.25 ksi) and 40 J (30 ft-lbs) for steels with specified minimum UTS :::: 450 N/mm2. BS 5500 requires impact testing weld metal of welding procedure test plates when the operating temperature is below 0°C (32°F), but does not req1:1ire impa.ct testing ofHAZ with multi-pass welding with weldmg heat inputs between 1 kJ/mm (25 kJ/in.) ~nd 5 kJ/mm (127 kJ/in.). Impact testing of product10n test plates is not required for welds in material less than 10 mm (0.394") thick. Stoomwezen requires two types of impact testing: • Quality test at 20°C (68°F) for all non-alloytd and low alloy steels. (The required impact test values are 31 J (23 ft-lbs) for carbon, C-Mn, and fine grain C-Mn steels with minimum specified UTS < 450 N/mm 2 and 27 J for UTS :::: 450 N/mm 2) • Extra testing at the impact test temperature determined from figures which relate assessment temperatures, thicknesses and impact test temperatures. (The required impact test values are 27 J) !mpact ~esting of procedure qualification test plates is required for service temperatures below 0°c. Production test plates need to be impact tested only when extra testing is required. Impact tests are required for base metal and weld metal, but not HAZ. (This is less restrictive than ASME Section VIII, which requires impact testing of weld metal and HAZ in carbon and low alloy steels whenever base metal needs to be impact tested). The Draft CEN. Standard lists three alternat·ive . meth od s to estabhsh impact test requirements· Method 1,. developed from past operating e~peri­ e~c~s, and is applicable to steels with specified mmimum UTS :s 460 N/mm 2 (66.7 ksi) and maximum ~hickness :s 30 mm (l.18") for as-welded constru~t10n and 60 mm (2.36") for PWHT construction. The impact test temperature is based on the minimurr: MDMT and an adjustment based on actual tensile stresses (% of maximum allowable). Th re~':ired impact test values are 27 J (20 ft-lbs~ mmimum average. Method 2. The base metal, weld metal and HAZ ~hall meet the required impact test values at the impact test temperature obtained from Fig's D 1 D.2, D.3 or D.4, based on the reference temper.at~r~ and thickness. Fig. D.1 and Fig. D.3 are for PWHT assemblies and Fig. D.2 and Fig. D.4 are for aswelded assemblies. The required impact test values are 27 J minimum average for materials with specified minimum UTS:::: 310 N/mm2 and 40 J (30 ft-lbs) for materials with specified minimum UTS > 310 N/mm 2 (45 ksi). Method 3 is a fracture mechanics analysis and may be used for materials not included in this CEN standard and cases not covered by Methods 1 and 2. Welding procedure qualification and production test plates must impact tested in thicknesses exceeding 12 mm (0.47''). Impact tests are required in base metal, weld metal and HAZ, except that no impact tests are required in the HAZ of low strength C and C-Mn steels (Steels in Groups 1.1and1.2). TheAD-Merkblatt Code requires impact testing of all carbon and low alloy steel base metals and welding procedure qualification and production test plates. For service temperatures - l0°C (l4°F) and above, the impact test temperatures and impact values ~or b~se metal ~nd weld metal are generally as specified m the apphcable material specifications. The HAZ shall meet 27 J (20 ft-lbs) ISO V-notch minimum average. For service temperatures below -10°C (14 °F), the base metal shall meet the impact test values in the applicable material specification (generally not less than 27 J (20 ft-lbs) minimum average) at the lowest service temperature. The weld metal shall meet 27 J minimum average and the HAZ 16 J (12 ft-lbs) at the lowest service temperature. In gener:f; most of the European codes and standards rev~ewed appear to have more severe toughness requir~ments thanASME Section VIII, particularly the thicker plates in as-welded condition. (See the comparison of impact test requirements for l" 1.5" and 2" ~hick plates in Section F.6 of this report)'. AS~E Sect10n VIII, however, have special toughness requirements (mils lateral expansion and also drop weight tests for some steels). In some cases the BS 550? and Stoomwezen codes do not require impact t~stmg of heat affected zones of procedure qualification and production test plates, whereas this is Evaluation of Design Margins, Section VIII 77 generally required for vessels built to ASME Section VIII, Division 1 and 2, when the base metal must be impact tested. Welded Joints Whereas various types of joints are permitted for ASME Section VIII Division 1 vessels, generally only Type 1 and Type 2 butt joints (for Category B joints only) can be used for Division 2 vessels. Type 1 butt joints are required for the high strength materials in Table AQT-1 in Division 2. Section VIII, Division 2 has very detailed requirements for the types of welded joints permitted, particularly for high strength steels. BS 5500 permits single-welded (with or without backing strips) or double-welded butt joints for main seams in vessels; however, single-welded butt joints are limited to 20 mm thickness, or less, depending on joint detail, for Construction Category 1 vessels. Also permanent backing strips must not be used for Construction Category 1 and 2 vessels. BS 5500 permits full penetration and partial penetration nozzle attachment welds and includes detailed information on nozzle attachment details and weld preps. Most nozzle attachment details are full penetration welds. Rules are also included for weld proximity. Appendix D includes suggestions for details to be used on vessels in low temperature service (below 0°C) to reduce the risk of brittle fracture. Stoomwezen permits one sided butt welds with backing strips left in place for girth welds for all materials. Also joggle joints are permitted under certain conditions. Full penetration and partial penetration welds are permitted for nozzle attachment welds, but Stoomwezen requires full penetration welds for all Hazards Category 3 welds and certain materials which require more care in processing and are more prone to cracking during their use (Stoomwezen Category C and CP materials). The Stoomwezen Code offers the greatest amount of design details amongst the pressure vessel codes studied in this report. Many of the weld details are adopted from recommendations by the International Institute of Welding. The Stoomwezen Code also includes a "suitability grading" table to help the designer select the appropriate design details for the intended service, loading conditions and the method of examination. Although most of the European pressure vessel codes and standards have provisions for use of various joint efficiencies in their designs (depending on the type of joints, welding processes and NDE), they are not as detailed as in ASME Section VIII, Division 1. Division 2 only uses joint efficiency of 1 since it requires a 100% NDE of welded joints. Inspection of Welded Joints ASME Section VIII, Division 1 permits full, spot and no radiographic examination for butt welded joints. Division 1 has no specific requirements for NDE of nozzle attachment welds for Part UCS 78 materials. Division 2 requires 100% RT of all Type 1 and 2 butt welds, UT of certain types of Category C and Category D joints, and 100% MT or PT of all nozzle and other attachment welds. The NDE requirements in the European codes depend on material, thickness, type of service, design stress and other considerations. For Construction Category 1 vessels, BS 5500 requires 100% RT or UT all shell butt welds, regardless of thickness, and 100% RT or UT of nozzle attachment welds when the thinnest part welded exceeds the specified thickness limit for a particular materials group. 100% MT or PT is required for all attachment welds. A lesser amount of NDE is permitted for Construction Category 2 vessels; however only certain materials are permitted for Construction Category 2 vessels and are limited to 40 mm thickness (30 mm for C-Mo steels). Only visual examination is required for Construction Category 3 vessels, but these are further restricted to the types of materials and thickness. The type and extent of the NDE in the ADMerkblatt Code depends on the type of material (material group), thickness and design stress. Generally 100% of RT or UT is required for all longitudinal shell butt welds and weld intersections and 100% or a lesser amount (e.g. 25% or 10%) for circumferential butt welds. More examination is required for higher strength materials and for alloyed steels (e.g. 100% RT or UT of all shell butt welds and weld intersections). The amount of RT or UT examination, however, can be reduced if the results of previous NDE (on similar vessels) reveals no serious deficiencies. (This is not an option for ASME Section VIII vessels). AD-Merkblatt requires MT or PT examination of all nozzle and other attachment welds. Nozzle welds also require RT or UT examination for nozzles exceeding certain size and connecting thickness. The extent of NDE requirements in the Stoomwezen Code depends on the hazard category, material group and thickness of the pressure part. Generally, more examination is required for higher hazard categories, more crack sensitive, and thicker material. Full (100%) examination is required for vessels of Category C or CP materials and for all Hazard Category 3 vessels. Hazard Category 2 vessels require either full or "extensive" examination depending on the material group and thickness. Hazard Category 1 require limited examination for Group A materials and either limited or extensive examination for Group B materials, depending on thickness. UT must be used in addition to RT for thicknesses exceeding 60 mm (30 mm for Category C and CP materials). UT must be used in place of RT for thicknesses exceeding 60 mm and for nozzle welds for all Category C and CP materials when the shell cutouts exceed 100 mm and nozzle wall thickness exceeds 20 mm. Surface examination (MT or PT) must be per- WRC Bulletin 435 formed on all material Category C and CP welds and on all welds of Hazard Category 3 vessels. The NDE requirements in the CEN draft Standard depend on the Risk Category, Testing Category of welded joints, group of steel, type of joint and maximum thickness. It has probably the most complicated set of requirements of all the European codes and standards discussed in this report. Longitudinal butt welds require 100% RT or UT for some Testing Categories and a lesser amount for others. The same applies to circumferential butt welds in vessel shells and to full penetration nozzle attachment welds with nozzle I.D. > 150 mm and thickness e > 16 mm. Butt welds which are RT or UT examined also require a certain amount of MT or PT examination. The amount or MT or PT of nozzle attachment welds again depends on the Testing Category, material group, the type of attachment weld and nozzle size/thickness. Some Testing Categories require 100% MT or PT, some less. A unique feature in the CEN standard is that in many cases the required amount of RT (or UT) and MT (or PT) can be reduced based on satisfactory prior experience (similar to AD-Merkblatt). This is not permitted in the ASME pressure vessel codes. The ASME Section VII, Division 2 appears to have the strictest and most well defined NDE requirements. The NDE requirements in BS 5500 appear to be comparable to those in Section VIII, Division 2. Also the AD-Merkblatt and Stoomwezen appear to have similar NDE requirements to Section VIII, Division 2; however the AD-Merkblatt Code permits a lesser amount of NDE in certain cases based on satisfactory prior experience, and the NDE requirements for nozzle welds in the Stoomwezen Code are not as well defined. It is difficult to compare the CEN draft Standard requirements to those for Section VIII vessels because of the many Testing Categories in the CEN Standard. The NDE requirements in t}ae European pressure vessel codes and standards generally are more extensive than those in ASME Section VIII, Division 1. Tolerances The out-of-roundness tolerance requirements in the European pressure vessel codes and standards appear to be similar to those for ASME Section VIII vessels; however, some of the European codes and standards have certain local tolerance requirements which are not included in Section VIII, Divisions 1 and 2. Some of these are: BS 5500 has limitations on local deviations (20° arc) for vessels subject to fatigue analysis and vessels constructed of steels with specified minimum UTS > 400 N/mm 2 . The Draft CEN Standard limits peaking at welded joint to 2% of a 20° are gage length. This standard also has detailed maximum offset limits for middle line alignments and surface alignment. Postweld Heat Treatment (PWHT) The European pressure vessel codes and standards generally have lower thiclmess limits for non-PWHT construction (at least for C and C-Mn steels) than ASME Section VIII. Also, the PWHT temperatures and hold times are generally lower (except for BS 5500, which has similar requirements to those in ASME Section VIII). Table 39 includes a summary of the thickness limitations for C and C-Mn steels. Third Party Inspection Generally more third party involvement is required by the European pressure vessel codes and standards, particularly the AD-Merkblatt and the Stoomwezen Codes and the CEN draft Standard thanASME Section VIII, Divisions 1and2. ' BS 5500. The Inspection Authority is commissioned by the Purchaser and is responsible for verifying the design and that the vessel has been constructed and tested in accordance with BS 5500. Its duties involve verification of materials, approval of welders and welding operators and welding procedure qualifications, examination of welds and NDE reports, heat treat records, and witnessing of pressure tests. (This is similar to the requirements in Section VIII, except verification of the design). The Authorized Inspector for AD-Merkblatt Code vessels has additional duties and responsibilities, such as NDE of materials (where needed), witnessing and supervision of NDE, conducting construction test (examination) and pressure test, and certification of vessels. The duties of the Notified Body (Inspection Authority) for vessels built to the Stoomwezen Code are similar to those for the ADMerkblatt Code. The CEN draft Standard requires the Notified Body (Inspection Authority) to verify that all parts of the vessel have been designed in accordance with the Table39 Code or Standard Maximum Thickness Without PWHT ASME VIII, Div. 1 1.5", providing welded joints over 1.25" thick are preheated 200°F ASME VIII, Div. 2 Same as for Division 1 35 mm (1.38"); 40 mm (1.57") provided the BS 5500 materials are impact tested to meet 27 J (20 ft-lbs) at -20°c (-4°F). 30 mm (1.18"); 38 mm with fine grain, AD-Merkbl~ impact tested material with minimum yield strength :s370 N/mm2 ; 50 mm provided material is fine grain steel with yield strength :s370 N/mm2, impact tested to meet 31 J at 0°C, welded assemblies with nozzles and attachments stress relieved, and brittle fracture assessment is performed. 32 mm (1.26"); 40 mm (1.57") provided Stoomwezen C :s 0.23%, CE :s 0.55, yield strength :s370 N/mm2 , base metal and WM meet 31 J at 0°C, and the vessel is subjected to hydrostatic test. 35 mm (1.38") >35 CEN draft Std mechanics Evaluation of Design Margins, Section VIII 79 CEN Pressure Vessel Standard and that the ves~el has been constructed and tested in accordance. ~1th this standard. The extent and level of the partici~a­ tion by the Notified Body depends on the vessel nsk category, testing category and the m?dule of conformity assessment. There is a reduced mvolvement by the Notified Body for a conformity assessment module involving a certified quality system. G. Recommendations for ASME Section VIII Division 1 and Division 2 Revisions a. Section VIII, Division 1 . Phase 1 of the PVRC report on design margms recommended that the design margin on ultimate tensile strength be reduced from 4.0 to 3.5, mostly on the basis of the improvements made to that document over the last three or four decades. The change is in the process of being implemented by the ASME Code Committee. (Code Case 2278, approved May 20, 1998; Code Case 2290, approved June .17, 199.8). Based on this Phase 2 study and comparisons with a number of European Codes, it is felt that f~1:t~er reduction in design margin for Section VIII, Div1s1on 1 is justifiable if a number of improvements are made to this Code. The value of design margin on ultimate tensile strength may be reduced ?elow down to approximately 3.0 for various service conditions, depending on how many of th~ .proposed improvements are adopted. The final revisions to be made and the value of the margin to use is, of c?urse, left to the discretion oftheASME Code Com1'.1ittee~. It is suggested that the following areas be mvestigated and, when deemed necessary, the Code .rules modified by the appropriateASME Code Committees for adopting a lower design margin: 3.? 1. Simplified fatigue analysis rules or fatigue exemption rules should be added. One of the weaknesses of the present rules is the fa~t that no guidelines are provided on what con~ti­ tutes cyclic service. Simple fatigue exemption rules, similar to those of ParagraphAD-160 of Division 2 would be helpful. If a vessel can not be exempted from fatigue analysis, then the analysis rules of Division 2 may be referenced. A number of cycles for exemption will have ~o be assumed (in the range of 500 to 2000 is recommended) and the design details and design rules made consistent for adequacy with that many cycles. 2. Design details should be reviewed and revised as needed. A number of details allowed by the existing rules, with a reduction in design margins, would provide a very low number of allowable cycles. Depending on the final design margin adopted and the nun_iber of cycles used for exemption. from ~ati~e analysis (item 1 above), the design det~Ils ':ill have to be modified. Some of the details with partial penetration and fillet attachment welds 80 will have to be disallowed. In particular, the details for nozzle attachments will have to be made more stringent. 3. The rules for design of formed heads should be updated. The existing rules are out of date and do not have a rational basis. These rules do not provide uniform safety margins over the range of parameters. Analytical methods are available to prepare a more rational set of rules. Test results and past experience can be used to justify analytical results. A proposed Code Case is being considered by the ASME Code Committee which will provide alternative design rules for formed heads). (Code Cases 2260 and 2261, approved May 20, 1998). 4. The rules for design of cone-cylinder junctions should be improved. The existing rules have evolved over a long time, without the basis for various rules having been documented. With today's analytical means, a more rational a:id consistent set of rules can be prepared. With reduced design margins, the satisfactory past experience will not be adequate justification for keeping the existing rules. 5. Design rules for reverse knuckles should be provided. Existing rules do not provide such rules and rules consistent with the reduced design margins are needed to assure that unsafe designs are not used. 6. Design rules for bolted flanges should be improved. The existing design rules an~ gasket factors have been in existence a long time and need to be updated. (At the time of this writing, the ASME Code Committee is ~ork­ ing on a revised set of rules for flange design). 7. Material toughness requirements should be reviewed and revised as needed. The existing toughness rules are based on fracture mechanics calculations, modified by successful past experience. For fracture mechanics c~lcula­ tions the stress in the region surroundmg the flaw is generally assumed as yield stress, to account for residual stresses in the as-welded condition. Therefore, the reduction in design margin may not provide significantly different results for the as-welded condition. However, a number of provisions are based on past experience with the existing design ma~gins. Such past experience will not be applicable with a significant reduction in design margins. Examples of such provisions are Par. UG-20 (which exempts certain carbon steels up to l" thickness from impact testing for minimum design metal temperatures as cold as -20°F) and paragraph UCS-66(c), (which allows exemption from impact testing for ferric steel standard flanges for temperatures as cold as -20°F). Other considerations are: a. The credit allowed by Figure UCS-66.1 for WRC Bulletin 435 vessels stressed to a value less tha~ the allowable stress will also have to be revised. b. The provisions of paragraph UCS-68(c), which allows a 30°F credit for PWHT (if not required by the Code) also needs refinement. Most European codes allow a credit for PWHT in their impact test exemption curves. It is recommended that also Section VIII, Division 1 includes provisions in the Code rules for impact testing to give benefit for any PWHT'd vessel. c. A general review of the toughness requirements in Division 1 for higher allowable stresses to insure the same margin of safety on brittle fracture for steels subject to Code toughness requirements, including the high strength Q & T steels in Part UHT. (See Sections C.4 and E.2.c of this report). 8. The required hydrostatic and pneumatic test pressures should be reviewed. With increased allowable design stresses, the stresses at the hydrostatic test pressure (1.5 times design pressure) may be excessive. The 1.5 factor is larger than most other codes and will probably have to be reduced. 9. Allowable stresses for over-pressure testing should be added. One shortcoming of the present code is the fact that no limits are specified for test stresses. 10. The joint efficiencies of paragraph UW-12 should be reviewed and, as needed, revised. With increased stress allowable design stresses, some of the joint types, specially fillet lap joints, may have to be disallowed or further limited. The types of joints allowed will influence the fatigue exemption rules. 11. The non-destructive examination requirements need to be reviewed and updated for higher allowable design stresses. For ex/. ample, Division 1, Part UCS presently has no specific requirements for NDE of nozzle attachment welds. With higher allowable design stresses, requirements need to be included in Division 1 for non-destructive examination of nozzle welds. 12. Further requirements on the qualifications of designers may have to be imposed. There is now a requirement in the Code "Foreword" that the designer be familiar with the computer programs used for design. This requirement may have to be expanded in general or for certain types of design. Considerations may include certification by a Professional Engineer, a listing of items to be included in the design specifications, and documentation of design calculations. 13. The role of the "third party" in reviewing design calculations may need to be expanded. Most European codes place the design respon- 14. 15. 16. 17. 18. 19. sibility with a party other than the vessel manufacturer. Under the existing ASME Code rules, the Authorized Inspector is only responsible to assure that design calculations exist. This responsibility may need to be expanded to some degree. This of course will have an effect on the qualification requirements of the inspectors. The requirements for environmental loads should be reviewed and inclusion of provisions for low probability accident conditions should be considered. The stresses allowed for conditions such as wind and earthquake may have to be revised, in light of reduced design margins. A number of European codes include certain provisions for accidental conditions. (In the draft CEN pressure vessel standard, these are called "exceptional design conditions.") For such conditions, lower design margins are justifiable. For such conditions, progressive plastic deformation, fatigue, and creep do not have to be considered in setting the allowable. Safe shutdown would normally be the criteria. The requirements for award ofASME's Certificates of Authorization may be reviewed and, if needed, expanded. Under the present arrangements, there are no specific requirements for design qualifications. With more sophisticated design rules, it will become more important for the Certificate holder to possess a certain degree of design capability. The rules for external pressure design should be updated. The present rules are old and do not cover all geometries and loadings of interest. A Code Case containing alternative buckling rules is being considered by the ASME Code Committee. (Code Case 2286, approved July 17, 1998). Design rules for penetrations, and particularly those for large penetrations should be reviewed. Code Case 2168 which provides an alternative to full area replacement rules and is based on existing allowable stresses will also have to be reviewed and revised. (At the time of this writing, PVRC is sponsoring a study by E.C. Rodabaugh to come up with design rules for large openings). Advantage shoullfl'be taken of the great deal of analytical work performed on shell intersections by PVRC and others. The desirability of allowing a lower design margin for materials with certain degree of ductility may be investigated. Most European codes differentiate between materials of low and high ductility. The design margin on ultimate tensile strength for ductile materials may be relatively low and still provide adequate safety margin. The desirability of basing the design margin Evaluation of Design Margins, Section VIII 81 on the consequences of failure may be investigated. Some of European codes differentiate among vessels in various types of service and classify vessels in different risk or hazard categories, depending on the toxicity of the fluids, pressure and volume of the vessel. The Stoomwezen Code and the proposed CEN rules provide such cla,ssifications. 20. Knuckle buckling for thin heads (D/t > 500) should be addressed. The present rules and the proposed Code Case for design of formed heads do not address buckling. PVRC is performing a project to come up with proposed rules. With higher allowable values of internal pressure, the knuckle buckling becomes more of a controlling factor. 21. External pressure curves should be extended into the creep range. The temperature limitations on these curves is a shortcoming of present code. PVRC has performed a great deal of work in this area which can be taken advantage of. 22. Although Section VIII, Division 1 includes requirements for out-of-roundness (UG-80) and edge alignment (UW-33), additional workmanship requirements may have to be added. With higher allowable design stresses considerations should be given to the following: a. Peaking at weld seams, local surface irregularities (flat spots) and verticality. (Examples of such tolerances are given in 4.2.3 of BS 5500 and in CEN draft Pressure Vessel Standard). b. Local imperfections at welds, such as weld undercut, excess gap at fillet welds, root concavity, etc. (examples of such limits are given in Table 5. 7(3) of the British Standard 5500). 23. Additional rules and cautions for vessels operating in the high temperature range may have to be added. Certain details and material combinations may not be advisable for high temperature operations. TheASME Code Committee and PVRC are working on such rules for high temperature design. With increased allowables, the effects of non-linearities will be more pronounced. 24. No consideration has been given in this report for increasing the allowable stresses in the creep range. The Code Committees, however, may want to make a review of the transition from time independent to time dependent properties. 25. Section VIII, Division 1 presently permits structural quality plates (SA-36 and SA-283) for pressure parts (except for lethal service) in thicknesses up through %11 with SA-6 delivery requirements and frequency of testing. With higher design stresses, these structural grade 82 26. 27. 28. 29. 30. 31. plates may have to be eliminated for use on pressure parts. SA-20 requires purchaser approval for repair welding of plates by the manufacturer. Some of the European pressure vessel codes require that all welded repairs on material by the manufacturer shall be indicated on test reports. With higher design stresses, this may also need to be included in the Division 1 requirements. The forming requirements in UCS-79 should be revised to include stress relieving of cold spun heads (unless the manufacturer can demonstrate that the material properties are not significantly altered). Higher design stresses are likely to increase distortion of bolted flanges. In addition to the stress criteria, there should be also a criteria for deformation and rotation of bolted flanges to prevent leakage. Presently there are some non-mandatory requirements in Appendix 5. More detailed mandatory requirements should be considered. Quality control requirements may have to be expanded. Appendix 10 of this Code includes a number ofrequirements that are to be incorporated into the Manufacturer's QC plan. Additional requirements are normally included in most manufacturers' QC plans which are not spelled out in this appendix. With lower design margins, an updating of this appendix will be advisable. The effect of lower design margins on allowables for shear stress and bearing stress should be investigated. Consideration should be given to providing values of yield strength and ultimate tensile strength for all Code allowed materials in Section II, Part D. Each Code section could then specify design factors to be applied to these values at design temperatures below the creep range. This will eliminate the need for generating and maintaining voluminous allowable stress tables. b. Section VIII, Division 2 Based on comparisons with the other codes and standards in this report and satisfactory past experience with vessels fabricated to these codes, it is concluded that the design margin on ultimate strength may be reduced from the present value of 3.0. The European codes have successfully used values as low as 2.35 on tensile strength and % on yield strength. A number of improvements in the Section VIII, Division 2 rules and additional requirements related to a lower design margin are recommended below. Depending on the improvements to be incorporated in the Code, it is felt that a design margin in the range of 2.3 to 2.4 on ultimate tensile strength may be justifiable. No reduction has been WRC Bulletin 435 proposed for the present design margin on yield strength. For materials with low ductility, a penalty factor may be warranted. The following is a list of some of the areas suggested for review by the ASME Code Committee, if a reduction in design margin is adopted: 1. Fatigue analysis rules should be improved. The fatigue rules of this code, which were developed in the 1960's, need to be updated. A great deal of data and design information has been generated since these rules were adopted. With higher stress allowables, it will be essential that a reliable fatigue analysis is performed for vessels in cyclic service. 2. Flange design rules should be improved. Comments made for flange design rules in Division 1 are applicable to this code as well. 3. All the European pressure vessel codes and standards reviewed as part of this study add the under-thickness tolerance for plates to the required thickness. With higher design stresses, consideration should be given in this practice for Division 2 also. 4. Design rules for formed heads should be improved. Although Division 2 rules are an improvement over those of Division 1, they still do not reflect today's understanding of behavior of such heads. 5. Design rules for cone-cylinder junctions should be improved. Analytical tools are available to develop more reliable rules than are presently in this Code. 6. External pressure design rules should be improved. These rules are presently the same as those of Division 1 and the comments made for Division 1 apply here also. 7. Notch toughness requirements should be reviewed and updated. Toughness rules of Division 2 are consistent with those of Division J· With increased allowables, these rules will need to be updated. Consideration should be given to: a. Notch toughness requirements for increased allowable stresses to provide the same (or greater) margin of safety on brittle fracture as presently in Division 2. (See Section E.2.c). Consideration should also be given to residual stresses when evaluating the toughness requirements, particularly for the high strength quenched and tempered (Q&T) steels in TableAQT-1. b. The quenched and tempered steels in Table AQT-1 have a high yield strength-to-tensile strength ratios and, therefore, have less tolerance for flaws and discontinuities. Additional NDE should be considered for welded joints when these steels used at lower design margins than are presently in Division 2. c. Section VIII, Division 2 presently gives no credit for PWHT in toughness requirements. Most European pressure vessel codes have less severe impact test requirements for stress relieved than for non-stress relieved weldments. Consideration should be given to including the credit for PWHT also in the Division 2 toughness rules. 8. The role of the "third party" in reviewing design calculations may need to be expanded. Most European codes place the design responsibility with a party other than the vessel manufacturer. Under the existing ASME Code rules, the Authorized Inspector is only responsible to assure that design calculations exist. This responsibility may need to be expanded. 9. Design rules for penetrations, and particularly large penetrations, should be reviewed. More up-to-date methods are available that could be taken advantage of. 10. Knuckle buckling for thin heads should be addressed. 11. The analysis methods of Appendix 4 should be updated. The rules were developed in the 1960's and were intended for analysis by the shells-of-revolution method. With today's widespread use of finite element analysis, the rules of this appendix are difficult to apply. The analysis rules should provide for finite element and other recent methods of analysis and specify acceptance criteria general enough to be applicable to various methods. 12. Section VIII, Division 2 includes requirements for shells (AF-130), heads (AF-135) and edge alignment (AF-140), however, additional tolerance and workmanship requirements may have to be considered. With higher allowable design stresses considerations should be given to the following (and possibly others): a. Peaking at weld seams, local surface irregularities (fiat spots) and verticality. (Examples of such tolerances are given in 4.2.3 of BS 5500 and in CEN draft Pressure Vessel Standard). b. Local imperfections at welds, such as weld undercut, excess gap at fillet welds, root co1'.cavity, etc. (examples of such limits are given in Table 5.7 (3) of the British Standard 5500). 13. Section VIII, Division 2 permits structural quality plates (SA-36 and SA-283 Gr. B & D) for pressure parts with SA-6 delivery requirements and frequency of testing. With higher design stresses, these structural grade plates may have to be eliminated for use on pressure parts. 14. SA-20 requires purchaser approval for repair welding of plates by the manufacturer. Some Evaluation of Design Margins, Section VIII 83 of the European pressure vessel codes require that all welded repairs on material by the manufacturer shall be indicated on test reports. With higher design stresses, this should also be considered for inclusion in the Division 2 requirements. 15. AF-111 in Division 2 permits any process for forming as long it does not unduly impair the mechanical properties of the material. AF-605 in Division 2 permits a maximum fiber elongation of 5% for the high strength materials in Table AQT-1 without stress relieving. Additional consideration should be given to loss of toughness due to cold forming for: a. Cold spun heads. (Cold spun heads should be stress relieved unless the manufacturer demonstrates that the material properties meet the specified requirements). b. Microalloyed steels. (See WRC Bulletin 322). 16. Higher design stresses are likely to increase distortion of bolted flanges. In addition to the stress criteria, there should be also a criteria for deformation and rotation of bolted flanges to prevent leakage. 17. The desirability of basing the design margin on the consequences of failure should be considered. Some European codes differentiate among vessels in various types of service and classify vessels in different risk or hazard categories, depending on the toxicity of the fluids, pressure and volume of the vessel. (See Appendix 3 on Stoomwezen Code and Appendix 4 on CEN draft Pressure Vessel Standard). 18. The addition of ultimate load design concept as a design tool should also be considered. In that case the acceptance criteria would need to be provided in Appendix 4 for that method. 19. As a long term project, an effort should be made to make the safety margins consistent for various components and for the different failure modes. With the existing rules, the true safety margins vary considerably from component to component and for different failure modes. It is recognized that this would be a very ambitious task and may not be totally feasible. H. Recommendations for Additional Studies The conclusion of this report is that the design margin on ultimate strength should be reduced for both Divisions 1 and 2 of Section VIII Code. However, a number of areas in these codes need to be reviewed and improved. The final values of the design factors and the extent of improvements made to these codes will of course depend on the judgment and collective experience of many Code Committees. To help the Code Committees make such decisions, additional studies by PVRC or other organizations 84 may be helpful. The following is a list of some of the areas that are recommended for additional studies: 1. Development of a more detailed list of recom- mendations for revisions to Division 1 (for a further reduction in the design margin below 3.5) and for Division 2. This phase 2 study has been fairly broad and the recommendations therein are mainly based on comparisons with other codes. As a result, the recommendations for revisions to the ASME Codes are not very specific. A follow up study, could assume definitive values of design margins and based on these values provide recommendations for specific revisions. 2. Notch toughness requirements should be reviewed and updated. With increased allowable stresses, the notch toughness requirements need to be increased to insure at least the same margin of safety on brittle fracture as presently in Division 1 and in Division 2. Consideration should be given to impact testing requirements both in the as-welded condition and in the PWHT condition. Particular attention should be given to the high strength Q & T steels in Table AQT-1, which have a high yield strength-to-tensile strength ratios and, therefore, have less tolerance for flaws and discontinuities. 3. A study of joint efficiency factors and NDE requirements for Division 1 to assure consistency of such rules. The joint efficiency used in design and the impact test requirements depend on the extent and effectiveness of NDE methods and acceptance criteria. With many revisions to these areas over the years, there may not be adequate consistency. 4. Development of design rules for openings, including large penetrations. A great deal of data and analytical results are available to develop more up-to-date and consistent set of rules for both Divisions. 5. Development of analysis rules and acceptance criteria. The rules of Appendix 4 of Division 2 should be updated. More up-to-date rules are needed to allow the more advanced analysis methods. Some European codes have adopted such rules. 6. Development of up-to-date fatigue rules. Technology exists for preparing rationale and up-todate fatigue design rules. 7. Review of criteria for high temperature design margins. This phase 2 report includes a comparison of rules for high temperature design in various codes. However, no recommendations are made for revisions. A further study could concentrate on this area and provide recommendations for updating the rules and design margins. 8. A detailed review of fabrication requirements WRC Bulletin 435 which are consistent with higher design stresses for both Division 1 and Division 2. 9. Review of Sections I, III and IV and B31 piping codes. To assure consistency amongst the various ASME documents, a review of various pressure vessel, boiler and piping codes under the jurisdiction of ASME may be advisable. This would be a long term project which may result in recommended revisions to these documents. 10. A qualitative cost-benefit evaluation of some of the more significant changes required to justify lower design margins. Bibliography ASME Code, Section VIII, Division 1, 1995 Ed. ASME Code, Section VIII, Division 2, 1995 Ed. ASME Code, Section II, Part D, 1995 Ed. ASME Code Cases: Boilers and Pressure Vessels 1995 Ed 5. British Standard 5500:1991. ' · 6. The German AD Merhblatt Pressure Vessel Code (1995). 7. The Netherlands Rules for Pressure Vessels (1994). an~ J!~g'~)~4 Draft Standard for Unfired Pressure Vessels (July 1995 1. 2. 3. 4. 9 9. Criteria of the ASME Boiler and Pressure Vessel Code for Design by Analysis in Section III and VIII, Division 2, ASME, New York, N.Y. (1968). . 10. ASME BC91-460, Revised Toughness Rules for Section VIII, Division 2. 11. Final R~P?~t on Evaluation of Design Margins for ASME Code Section VIII, D1v1S1on 1, K. Mokhtarian and E. Upitis. (Report No. 1 ofWRC Bulletin No. 435) 12. WRC Bulletin ~5, April 1~64, PVR9 Interpretive Report of Pressure Vessel Research, S~ction 1-Design Considerations, B. F. Langer. 13. WRC Bulletin 101, November 1964, PVRC Interpretive Report on Pressure Vessel Research, Section 2-Materials Considerations, J. H. Gross. 14. WRC Bulletin 175, August 1972, PVRC Recommendations on Toughness Requirements for Ferri tic Materials. 15. WRq Bulletin ~35, .August 1988, A Review of Area Replacement Rules for Pipe Com.iect10ns m Pressure Vessels and Piping, E. C. Rodabaugh. 16. WRC Bulletin .406, November 1995, Proposed Rules for Determining Allowable Compressive Stresses for Cylinders, Cones, Spheres and Formed Heads, C. D. Miller and K Mokhtarian. 17. Design Stress Basis for Pressure Vessels, B. F. Langer (The William M. Murray Lecture, 1970), Experimental Mechanics, January, 1971. .18. ASME 71-PVP-49, Experimental Effort on Bursting of Constrained Discs as Related to the Effective Utilization ofYield Strength, W. E. Cooper, E. H. Kottcamp, and G. A. Spiering. 19. Second International Conference on Pressure Vessel Technology Paper II-77, October, 1973, Effect of Strain of Hardening on Bursting Behavior ofPressure Vessels, C. P. Royer, S. T. Rolfe and J. T. Easley. 20. ASME 74-Mat-1, ~ournal of Engineering Materials and Technology, Effect of Strain-Hardemng Exponent and Strain Concentrations on the Bursting Beha~ior of Pressure Vessels, C. P. Royer and S. T. Rolfe. 21. International Journal ofPressure Vessel and Piping, Vol. 22, 1986, pp 147-159, Pressure Testing of Large Scale Torispherical Heads Subject to Knuckle Buckling. 22. Journa! of Pressure vessel Technology, Vol. 110, August 1988, pp 226-233, Statistics of Pressure Vessel and Piping Failures, S. H. Bush. 23. Journa! of Pr:ess.ure Vess~l Technology, Vol. 110, Nov. 1988, pp 430-443, Des~gn Critena for Boilers and Pressure Vessels in the U.S.A., M. D. Bernstein. 24. Structural Analysis and Design of Process Equipment, Second Ed., M. H. Jawad and J. R. Farr, 1989, John Wiley & Sons, Inc. 25. Summary of Design Analysis Factors Inherent in the Established Failure Modes of The ASME Boiler and Pressure Vessel Code ASME ' Subgroup Design Analysis (SCD) report, February, 1989. 26. PD 6493: 1991, Guidance on Methods for Assessing the Acceptability of Flaws in Fusion Welded Structures. 27. Brittle Fracture Margins in ASME Code, Section VIII, Division 1, S. Yukawa, Jan. 7, 1994 (Prepared for MPC Program on Fitness-for Service). 28. SME PVP, Vol. 277, 1994, p. 79-88, Safety Margins of Pressure Vess.els When Designed by Primary Stress Limits, A. Kalnins and D. P. Updike. 29. {'!?ME PVP, Vol. ?77, 19.94, p. 89-94, Burst by Tensile Plastic Insta.bihty of Vessels With Tonspherical Heads, D. P. Updike and A. Kalmns. 3~. ONR-AISI Agreeme~t No. N00014-94-2-0-002, Effect ofYield-Tensile ~at10 on Structural Behavior-High Performance Steels for Bridge Construction, R. L. Brockenbrough & Associates, Inc. (Expanded Draft Final Report), June 6, 1995. 31. ASME PVP, Vol. 315, 1995, p. 253-284, Hydro-test of 'I\vo Retired Pressure Vessels With Local Thin Areas, L. M. Connelly. 32. The Draft GEN Standard for Unfired Pressure Vessels, Seminar Papers, 25th September, 1996, IMechE, London, England. 33. D~velopment of a New Pressure Vessel System in the European Union, F. Osweiller, Sept. 20, 1996. (ASME B & PV Code: Main Committee Meeting). 34. W. D. Doty, Committee correspondence to G. Karcher September 21 1997. , , 35. WRC Bullet~n 265! February 1981, Interpretive Report on Small Scale Test Correlat10ns with Kro Data, R. Roberts and C. Newton. 36. API 579 Draft, Issue 6, February 15, 1997. 37. ASME Code, Section IX, 1995 Ed. 38. ASME Code Case 2235. 39. A Study of StreifS Relief Heat 1}-eatment for 89 kg I mm2 Class High !?trength Steels, M. fiiromatsu, Y. Kasamatsu and K Horikawa, Proceedings of an Internat10nal Conference on Welding of HSLA (Microalloyed) Structural Steels? 9-12 November, 1976, Rome, Italy, p. 503-529. 40. P~C Project.J'!o: 97-2, Evaluation of Design Margins for ASME Code, Section VII~, D!Vls10ns 1 and 2 Phase 2 Studies, including Appendices 1-4, Final Report, November 1997. 41. Order on Pressure Vessels, Gas Pressure Vessels and Filling Plants (Germany), November 1992. 41. European Pressure Equipment Directive 97/23/EC, 29 May 1997. Evaluation of Design Margins, Section VIII 85