CONFERENCE PROCEEDINGS 1

1

Eperc International Conference

Pressure Equipment Innovation and Safety – Vol. 2

Roma, 1-3 aprile 2019

1

Eperc International Conference

Pressure Equipment Innovation and Safety

vol. 2

Roma, 1-3 aprile 2019

Department of technological innovations and safety of plants, products and anthropic settlements (Dit)

Editorial board

Daniela Gaetana Cogliani Inail Dit, Italy

Technical editor

Andrea Tonti Inail Dit, Italy

Scientific Committee

Claude Faidy Ahmed Shibli Guy Baylac Gerard Lackner Fernando Lidonnici Piet Verbesselt John Wintle Roger Hurst Stefan Holmstrom Dinesh Chand Sharma Augusto Di Gianfrancesco Egidio Zanin

Alistair Klein Bart Teerlink Isamu Nonaka Kouichi Maruyama Yasuhiro Yamazaki Giancarlo Gobbi Luana Campanile

Eperc Chairman, France Etd, United Kingdom

Eperc Technical Advisor, France Tuv, Austria

Sant'Ambrogio, Italy Consultant, Belgium

Consultant, United Kingdom Consultant, United Kingdom Dg Jrc Ec, Finland

Sesei, India

Eccc Chairman, Italy

Rina Consulting – Csm, Italy Astm, United States

Sirris, Belgium

Tohoku University, Japan Tohoku University, Japan Chiba University, Japan Asme, United States

Scientific Committee Technical Secretariat, Italy

info

Inail - Department of technological innovations

and safety of plants, products and anthropic settlements via Roberto Ferruzzi, 38/40 - 00143 Roma

dit@inail.it www.inail.it

© 2019 Inail Isbn

The authors are fully responsible for the opinions expressed in the publication, which must not be understood to be official opinions of Inail.

Publications are distributed free of charge and therefore sell and reproduction via any means are prohibited. Citation is only permitted with reference to the source.

978-88-7484-187-5

Technical Plenary Session

P. Smith, D. Nash, S. Earland

Pressure Systems Activity in a post-Brexit UK ... 1

Design and fabrication Session

E. Becherini, S. Milani, A. Muratore, V. Nastasi

Comparative pressure vessel calculation of seismic effect: European versus American concepts and codes ... 5 A. Saccocci, P.E. Di Nunzio

Metallurgical design of high-toughness, high-strength steels for 300 bar gas cylinders applications ... 15 C. Faidy

Nonlinear Analysis in Pressure Vessel Design Codes Recommendations for Codified Rules Improvements ... 24

Fitness for Service

C. Faidy

French Fitness for Service – Codes Status and Open Points ... 37 A. Marino, M. Ciucci, L. Barbieri Vita, O. Palermo

Integrated Smart Approach to Seismic Risks Management In Process Plants ... 51

Non-Destructive Examination - In Service Inspection and Operation Session

Biancolini M.E., Brutti C., Zanini A., Salvini P.

Acoustic Emission data fractal analysis for structural integrity assessment of pressure equipment ... 60 M.E. Biancolini, C. Brutti, A. Chiappa, P. Salvini

Artificial pre-cracking of tanks test samples for AE detection tuning ... 69 G. Merckling, L. Casiraghi

Microstructural Parameters and Creep Exposure for 9Cr Steels. A Tentative Quantitative Correlation Based on “Metallic” Replication ... 78 A. Salvo, A. Staffolani, D. Benini, A. Corsi, G. Merckling

Residual Life Assessment of Metallic Materials by an Innovative Non-Destructive

Metallographic Test ... 88

1

Pressure Systems Activity in a post-Brexit UK

P Smith*, D Nash**, S Earland***

* Wood plc.

** University of Strathclyde.

*** Earland Engineering Ltd.

Summary

The UK was scheduled to leave the European Community at the end of March 2019 but the deadline has been extended and the situation regarding the UK exit from the EU remains in a state of uncertainty. However the work of pressure systems transcends the political, economic and freedom of movement challenges and there is a requirement to continue to strive towards excellence in design codes and standards and to maintain equivalent CE conformity in manufacturing.

The Institution of Mechanical Engineers Pressure Systems Group and the British Standards Institution’s Pressure Vessel Engineering committees all support this agenda and are working towards maintaining the UK’s engagement in a global pressure equipment sector.

This paper addresses the key challenges facing the UK pressure systems community and points towards the UK’s continuing engagement with CEN and highlights future developments in certification and future code research and development.

1. Pressure Systems Group

The authors of this paper are all members of the Pressure Systems Group Board at the IMechE.

The Pressure Systems Group (PSG) is a community of key workers from many industries.

Events include seminars, conferences and courses (in conjunction with Strathclyde University).

The PSG also presents the Donald Julius Groen Prize. Past winners include Clause Faidy (EPERC Chairman), Fernando Lidonnici (Convenor of CEN/TC 54 WG 53 – Design Methods) and Guy Baylac (Convenor of CEN/TC 54 WG 59 – Creep).

2. Legislation

If the UK leaves the EU without a withdrawal agreement then the following changes will happen as soon as the UK leaves the EU.

If the UK leaves with a withdrawal agreement then the changes will be subject to the terms of any subsequent trade agreement.

The situation in Westminster changes hourly and so the arrangements outlined in this paper are obviously subject to change.

3. Essential Safety Requirements

The essential safety requirements applying to pressure equipment placed on the market in the UK will not change, except as noted below.

PED 2014/68/EU [1] was implemented into UK law by The Pressure Equipment (Safety) Regulations 2016 (SI 2016 No.1105) [2].

2 These regulations will be amended to remove references to the EU and to replace the CE mark with a new UKCA mark – see Figure 1.

The Essential Safety Requirements (ESRs) will be unchanged except for:

The removal of the reference to materials covered by a European approval for materials – see ESR 4.2(b).

The removal of the reference to material manufacturers that have an appropriate quality- assurance system, certified by a competent body established within the Union – see ESR 4.3, third paragraph. This has led to some confusion regarding the status in the UK of certificates issued by EU material manufacturers going forward.

If the UK leaves the EU without a withdrawal agreement then the Pressure Equipment (Safety) Regulations, along with many others, will be amended by a single bill which is currently before the UK Parliament.

4. Conformity Assessment – New UK Arrangements

A new UK framework for conformity assessment will come into effect when the UK leaves the EU.

The underlying rules and regulations setting out the requirements for conformity assessment for each product will stay the same.

A new UK conformity marking, the UKCA mark, will indicate that a product complies with UK regulations and can be placed on the UK market.

Figure 1. The UKCA Mark

5. Using CE marking in the UK

Pressure equipment that meets the relevant EU regulatory requirements and bears the CE mark can still be placed on the UK market for a limited time period after the UK leaves the EU.

The UK government will give notice before this period ends.

This arrangement applies whether the CE marking is used after the manufacturer has self- declared it, or after an EU-recognised conformity assessment body has assessed its conformity.

6. Notified Bodies

UK Notified Bodies will automatically become UK Approved Bodies.

This represents their new status in UK law after the UK leaves the EU, but their role in conformity assessment will not change.

UK Approved Bodies will not be recognised as Notified Bodies by the EU. UK manufacturers will need to use an EU Notified Body for pressure equipment that is to be exported to an EU country.

EU notified bodies are listed on the New Approach Notified and Designated Organisations (NANDO) database. After the UK leaves the EU details of UK approved bodies will be available on a similar UK government database which will be publicised ahead of EU exit.

3

7. Standards

The role of standards in the UK regulatory framework for manufactured goods will not change when the UK leaves the EU.

The current EU harmonised standards will be carried across as UK designated standards to maintain a single standards model between the UK and the EU.

UK designated standards will provide a presumption of conformity with the applicable UK regulations, in the same way that EU harmonised standards provide a presumption of conformity with the applicable EU directives.

CEN, the European Committee for Standardization, is an association that brings together the National Standardization Bodies of 34 European countries, including 6 that are not members of the EU.

The British Standards Institution (BSI) will remain a member of CEN after the UK leaves the EU.

The decisions of the CEN and CENELEC General Assemblies in November 2018 set in place a plan that secures BSI’s full membership of these organizations when the UK leaves the EU.

In terms of BSI’s continued membership a transition period for the CEN and CENELEC statutes will begin on the date of the UK’s effective withdrawal from the EU, until the end of 2020.

BSI experts will still be entitled to participate in CEN and CENELEC technical committees and BSI will continue to enjoy full voting rights in both organizations’ decision-making bodies.

8. Pressure Vessel Standards

The UK standard for unfired pressure vessels is, and will remain, BS EN 13445 [3].

Under the European Committee for Standardisation (CEN) rules BSI was obliged to withdraw BS 5500 [4] when the European Standard was published in 2002.

EN 13445 was not complete, and due to demands from UK industry BS 5500 was reissued as a published document PD 5500 [5], with equal content, validity and application, but without the status of a "national standard".

There are no plans to withdraw PD 5500 at the present time.

BSI has a number of committees supporting PD 5500 – and two, PVE/1 Strategy and Policy and PVE/1/15 Design Methods are actively developing the code.

A number of new topics are being researched and the outputs of these can be incorporated into PD 5500 in a quicker and more streamlined process than for CEN standards.

Recent developments in PD 5500 include the nickel, titanium and copper supplements, and similar developments are being incorporated into EN13445.

There are some concerns over capacity for future developments on an on-going basis.

9. Conclusions

The UK relationship with the EU remains in a state of flux.

Whatever the outcomes of the political process, the UK pressure systems community will have a route to ensure compliance with the PED, and EU manufacturers will have a route for compliance with UK regulations.

The UK has arrangements for an equivalent to the CE mark – the UKCA mark – which can be implemented if necessary.

BSI remains a member of CEN and will continue to work during transition until 2020 and beyond.

4 The UK and its experts will still participate in technical committees and working groups for the development of CEN standards.

The UK will continue to actively undertake research and development in support of PD 5500.

The Pressure Systems Group of the IMechE is the main UK focus for mutual professional relationships e.g. with ASME, JSME, ICPVT, EPERC etc.

The Pressure Systems Group will continue to serve its members in training and developing its engineers and mounting seminars and workshops showcasing latest developments and technologies.

10. References

[1] EUROPEAN COMMUNITIES. Directive, 2014/68/EU (PED) of the European Parliament and of the Council of 15 May 2014 on the harmonisation of the laws of the Member States relating to the making available on the market of pressure equipment (Official Journal of the European Union L 189/164, 27.6.2014).

[2] GREAT BRITAIN. The Pressure Equipment (Safety) Regulations 2016. SI 1105.

London: The Stationery Office.

[3] EN 13445:2014+A3:2017 (all parts), Unfired pressure vessels.

[4] BS 5500:1994, Specification for unfired fusion welded pressure vessels (withdrawn).

[5] PD 5500:2018+A1:2018, Specification for unfired fusion welded pressure vessels.

5

Comparative pressure vessel calculation of seismic effect: European versus American concepts and codes

E. Becherini *, S. Milani *, A. Muratore**; V. Nastasi***

* Ener Consulting

** Inail UOT Como

*** Inail UOT Palermo

Summary

The aim of this paper is to show the “modus operandi” of seismic pressure vessel design and to compare European and USA main standards, highlighting the principal differences between these approaches. Indeed we remember that, if loads different from pressure are neglected or applied improperly to the equipment, the vessels could become unsafe, since they can be partially or completely undersized.

The mechanical calculations for a device under pressure can be usually well-defined starting from procedures based on mechanical calculation codes. However, their integration with building and seismic codes often may result difficult during the design phases.

Therefore, the excursus will be the following:

 To understand when the loads due to the earthquake have to be applied;

 To understand how the main regulations for the mechanical design deal with the building codes and loads;

 To show some examples of evaluation of seismic loads using different regulations.

1. Introduction

Referring to the European and the international contexts, it is extremely important to consider that both the 2014/68/UE Directive (Pressure Equipment Directive - PED) [1] and the ASME Boiler and Pressure Vessel Code [2] impose that the design of a pressure equipment has to take into account, if necessary, loads different from pressure, such as the seismic loads.

The main equipment calculation codes, such as the PED harmonized standard EN 13445 [3], the American code ASME Section VIII Division 1 [4] “Rules for Construction of Pressure Vessels” and ASME VIII Division 2 [5], require to consider a series of loads, reported in the following Tables 1, 2, 3 and 4.

6 Table 1 – Loads to consider according to EN 13445

Table 2 – Loads to consider according to ASME VIII Div.1

7 Table 3 – Loads to consider in ASME VIII Div.2 in case of DBF

Table 4 - Loads to consider in ASME VIII Div.2 in case of DBA

Contractors often send to the Manufacturers detailed specifications as far as to consider loads other than pressure for calculation but, in many other cases, very little information is given to the designer.

It is clear that only a careful knowledge of the installation site and therefore of all the boundary conditions will lead to valid design calculations.

8 It is important to note that all previous tables require to consider seismic loads.

2. Treatment of the codes for calculating loads other than pressure

European EN 13445 and ASME Codes, with some difference, subdivide the loads into:

 Dead Loads / Permanent Action – Gk: loads due to the weight of the structure, including the parts attached to it (externally and / or internally), etc.;

 Live Loads / Variable Action - Qk: loads which are variable over time such as snow, wind, weight of the maintenance personnel with their tools, vibrations, etc.;

 Exceptional Loads - A: due to explosions, fires, impacts, etc.;

 Seismic Action - EE

Therefore, a number of known or predictable actions could coexist in a pressure equipment: Design pressure, Design temperature, External pressure, Dead Loads, Snow load, Wind load, etc.

All these loads must be determined and then combined as request by the codes with a predefined “weight”.

However only some calculation standards indicate how to combine the loads: for example, EN 13445 (for tall vertical vessel) and ASME VIII Div.2.

3. Building Codes and Pressure equipment standards

It is our opinion that in Italy as well as in other European countries, the field of application between pressure equipment codes and building codes is not well defined, for loads other than those typical of pressure and structural analysis. This situation may represent a dangerous overlay of engineering laws (building laws) and pressure equipment standards.

Indeed, the building codes, as for example the NTC 2018 [6], that is an Italian law, as well the Eurocodes, may be surely a valid aid to found the seismic, wind and snow loads but, they do not deal with the design of pressure equipment and their ancillaries as skirts, saddles, legs, etc.(as it is often practically required).

Unfortunately, NTC 2018 in Par. 7.2.4. “Plants design criteria“ as well law D.M. 02/02/1974

“Measures for buildings with special requirements for earthquake zones” [7] could create a dangerous overlap, because building codes have been written for a different purpose: in fact building codes don’t admit pressure vessel materials and temperature and pressure loads are not considered.

The European Regulation CPR 305/11 with the standard EN 1090/1 “Execution of steel structures and aluminum structures” [8] can represent another critical issue: because there may be another overlap with the Pressure equipment codes. This standard in fact by means of its FAQ 31 [9] states that saddles, skirts, legs, etc., not calculated by means of EN 13445, must be CE marked according to EN 1090/1.

Moreover, the new European approach on building codes is based on the limit states calculation and prohibits de facto the classic approach with the allowable stress method.

This situation is an unsolved problem with Civil European laws and with Civil National Authorities.

In USA, the situation between building and pressure equipment codes, as well for tank codes, is more clearly defined.

In fact, American building code ASCE 7-16 Standard [10] “Loads for Building and Other structures” and the ASME Boiler and Pressure Vessel Code are, for non-pressure loads, like seismic or wind loads, rational and well-integrated codes.

Moreover, the USA approach permits to the designer to use both LRFD (Load and Resistance Factor Design – Limit State) and ASD (Allowable Stress Design) methods for ancillary parts of pressure equipment: of course, allowable stress method ASD is more manageable for pressure equipment designers.

9

4. European EN 13445 approach

EN 13445 is a standard that provides rules for the design, fabrication, and inspection of pressure vessels and is harmonized to the Pressure Equipment Directive (2014/68/EU or

"PED"). EN 13445 deals with loads different from pressure, in the 16^th chapter “Additional non-pressure loads”.

The rules proposed in this chapter are accurate, since for each kind of load, they take also into account the effect due to pressure and they permit to verify if the mechanical stresses or the local and global stress / buckling are allowable; but at the moment, they do not give definitive load combinations for seismic and wind loads on global analysis.

An important new revision of the standard is Clause 22 “Static analysis of tall vertical vessels on skirts” which explains how to combine the different loads.

Clause 22 is linked to the pertinent building Eurocodes to define the design value.

Load combinations for tall vertical vessel imposed by EN 13445 are:

Table 5 – Load combinations for tall vertical vessel

Surely the new clause 22 has given to the designers a very important contribute but we hope that in future EN 13445 will use this new approach also for not tall vertical vessels.

An alternative and innovative approach, suggested by the EN 13445 ANNEX B “Design by Analysis – Direct Route”, introduces the concept of Limit State. This concept could be found into the previously mentioned building codes.

Annex B introduces the concepts of Ultimate Limit State and Serviceability Limit State.

The first one defined as the state that could be associated to collapse of the structure, while the second one is defined as a state beyond which the service criteria specified for the component are no longer met.

Some examples of Ultimate Limit State are failure by gross plastic deformation, rupture caused by fatigue, collapse caused by instability of the vessel or part of it, loss of equilibrium of the vessel or any part of it, considered as a rigid body, by overturning or displacement and leakage which affects safety.

Examples of Service Limit State are deformations or deflections which adversely affect the use of the vessel (including the proper functioning of machines or services), or cause damage to structural or non-structural elements, leakages which affect efficient use of the vessel but do not compromise the safety nor cause an unacceptable environmental hazard.

Summarizing, it can be stated that this methodology of calculation shifts the reasoning from stresses to loads. In particular, if in the allowable stress approach the equivalent stresses are evaluated (Tresca or Von Mises) and compared with the allowable stresses, in the Limit state calculation, a specific combination of loads is compared with the maximum admissible load, that is function of the material used.

In other words, the design actions (Ad) are evaluated as:

10 Ad = ƔA A

where A is the characteristic value of the action and ƔA is the partial load safety factor.

These actions have to be compared with RMd , which is the design value of the material:

RMd = RM / ƔR where RM is the characteristic value of the material and Ɣ the partial material safety factor.

The value of RM is chosen according to material standard codes (ReH, Rp0.2/T, Rp1.0/T;

Rm/T, etc.).

Paragraph B.5.2.1 imposes a check that has to be done using this methodology (Table 6), whereas paragraph B.5.2.2. explains the steps to be followed to perform the evaluation (Table 7).

Table 6 – Analysis to satisfy for Annex B

Table 7 – Checks to satisfy for Annex B

The single action, for different loads, could have positive or negative effects. For instance, in vertical vessel under the action of the wind, the pressure decreases the axial stress of compression into the shell downwind and therefore the effect of the wind is positive in this case. It follows that for each design check (GPD, PD, etc.), different partial safety coeficients exist, depending on the type (positive or negative) of the effect.

An insolved problem still exits, since the regulations do not specify how to combine the loads for the Ultimate Limit State and Serviceability Limit State analysis.

For these considerations at the moment, we refer to the SAFAP 2014 pubblication

“Verifica delle attrezzature a pressione per carichi diversi dalla pressione: Vento e Sisma” [11]

Considering many combinations of loads implies the use of finite elements softwares.

The softwares should be technical and high- quality since the regulations recommend the type of elements to be used for the modelling (Shell, Beam, Brick, etc.). Finally, these complex softwares should be used only by specialists, otherwise the results of the analysis could be wrong.

11

5. The approach of the American regulations ASME VIII Division 1 e Division 2

ASME VIII Div.1 admits only DBF approach for non-pressure loads, refers to consolidated engineering methods (Zick Method, Brownell & Young, and other), although without explicitly mentioning them. These methods allow to evaluate stresses due to loads other thanpressure, but only if they are “as safe as” according to what reported on the U-2 (G).

Indeed, each stress, on parts of the equipment, has to be lower than the allowable stress as evaluated by ASME VIII Div.1. Generally the adopted method for loads application is the equivalent static, and the seismic loads are taken from ASCE 7-16, UBC [12] or other National Building Codes as ASME VIII Div.1.

ASME VIII Div.2, as reported below, allows two kinds of analysis:

 Design by Formulae (Part 4);

 Design By Analysis (Part 5).

Neglecting Part 4 which uses a design by formulae approach, the 5^th part is related to a calculation that enables to analyze the following cases:

 Protection against Plastic Collapse

 Protection against local failure

 Protection against Collapse from Buckling

 Protection against failure from cyclic loading

ASME VIII Div.2 specifies also how loads should be considered and how they should be combined for an Elastic or Elastic-Plastic Analysis.

This methodology is similar to EN13445 annex B approach, but here we can found the load combination, that in EN code is missing.

Also for ASME VIII Div.2 the seismic loads are taken from ASCE 7, UBC as for ASME VIII Div.1.

6. Design example of European approach versus USA approach

Here below we show two different case studies of real job for seismic calculation of vertical vessel for two different leader customers, one in north Europe and another one in California (USA).

 Vertical Exchanger installed in north Europe

The vessel has been designed for a temperature Ts of 450 °C and a design pressure of 13 bar-g.

Customer specification requirements were:

Certification of the equipment: 2014/68/EU – PED [1]

Certification of the skirt: EN 1090 (CPR 305/11)

Static calculation: Eurocodes 3

Admitted Pressure equipment code: EN 13445

Adimitted wind and seismic standard: Zone 1 (DIN 1055-4)

Seismic factor : Zone 0 (DIN 4149)

For the mechanical calculation of the equipment with EN 13445, regarding seismic and wind calculation, the customer required that at least the skirt calculation had to be performed in full compliance with Eurocodes, with Limit State verification, with all loads combinations and wind and seismic loads applied from different directions.

This procedure approach imposed to the designer an interpetation to match EN 13445 and Eurocodes with a lot of load combinations, with loads as pressure and temperature which are not contemplated, as well as low alloy steels, by the Eurocodes. The same

12 problems occurred for the EN 1090 certification because the low alloy steel are not in compliance with the standard who deals only with carpentery steel work.

Here below some examples of imposed loads on FEM model for a design check with Eurocodes:

Figure 1 – Example of calculation (EN 13445 + Eurocode 3)

As below written, a big problem has been to perform the check required by Eurocodes in terms of loads and not in term of stress.

 Vertical vessel installed in USA

The following example is a seismic analysis required in the calculation of a reactor installed in California (USA). The vessel has been designed for TS 427°C and PS 70 bar- g. In this example is shown a FEM model to verify the region of junction shell – skirt. The vessel has been certified ASME VIII Div. 2 (U2 Stamp).

The customer specification requirements for seismic analysis were:

Seismic Load Code: ASCE 7-16

Site Class: D

Spectral response param at short period, Ss: 1.768 Spectral response param at 1 second, S1: 0.631

Importance Factor, Ie: 1.25

Long Transition Period, TL: 8 seconds

Response modification coefficient, R: 3

Previous parameters have been used in ASCE 7-16 to find loads and earthquake acceleration.

The vessel has been designed according to ASME VIII Div.2 class 2, and the load combinations have been defined according to Table 5.5 of ASME VIII Div.2 (see table 8).

13 Table 8 – ASME VIII DIV.2 Part. 5 – Load case combination

Table 8 includes all applicable loads on a vessel (design, enviromental, live loads etc), and permits a full strength check i in vessel (or column).

It is important to highlight, again, that ASCE 7-16 is interconnected in a clear manner to ASME codes for pressure equipment.

Here below some examples of imposed load on FEM (a sub model is used) and skirt results:

Figure 2 – Example of calculation – ASCE 7-16 + ASME VIII Div.2

7. Conclusion

In this paper we explain some problems which may result with application of Pressure Equipment mechanical Codes and Nationals Building Codes for earthquake application loads.

For us, USA approach appear more linear and integrated than Europe approach.

It is desirable that in Europe, for the design of new Equipment, Nationals Building Codes will be more clearly connected to Pressure Equipment mechanical Codes for pressure equipment subject to earthquakes.

Moreover, we hope, that “more clearly connected approach” of Nationals Building Codes and Pressure Equipment mechanical Codes, will be useful for seismic verification of the existing plants. For example, it is required, by Directive 2012/18/EU (Seveso Directive),

14 but actually in Europe a dedicate standard verification (or guideline) for old pressure equipments subject to earthquakes doesn’t exist.

This article was written to improve the safety level of pressure equipment subject to earthquake and to facilitate the work of designers and inspectors.

References

[1] Directive 2014/68/EU of the European Parliament and of the Council - Pressure Equipment Directive, Bruxelles, 2014

[2] Boiler and Pressure Vessel Code, The American Society of Mechanical Engineers, New York, 2017

[3] EN 13445 Unfired pressure vessels, CEN

[4] ASME VIII Division 1 - Rules for Construction of Pressure Vessels - The American Society of Mechanical Engineers, New York, 2017

[5] ASME VIII Division 2 – Alternative Rules for Construction of Pressure Vessels - The American Society of Mechanical Engineers, New York, 2017

[6] Decreto 17 gennaio 2018 – Norme Tecniche per le Costruzioni – Gazzetta Ufficiale Italiana

[7] Legge 2 Febbraio 1974 – Provvedimenti per le Costruzioni con particolari prescrizioni per le zone sismiche - Gazzetta Ufficiale Italiana

[8] EN 1090: 2009 - Execution of steel structures and aluminum structures - Requirements for conformity assessment of structural components - CEN

[9] FAQ 31 – Frequently asked question on CPR – European Commission; 2017

[10] ASCE 7-16 – Minimum Design Loads and Associated Criteria for Buildings and Other Structure; American Society of Civil Engineer; 2016

[11] Atti Safap 2014 – Verifica delle attrezzature a pressione per carichi diversi dalla pressione: vento e sisma” – E. Becherini, F. Zichichi; 2014

[12] UBC 97 – Uniform Building Code – International Code Council; 1997

[13] Antonio Cirillo, Il Manuale dell’Ingegnere per il Calcolo Strutturale Tomo 1 e Tomo 2, Sistemi Editoriali Se, Napoli, 2012

[14] Donatello Annaratone, Pressure Vessel Design, Springers, Berlino, 2007

[15] Fernando Lidonnici, Progettazione apparecchi a pressione e scambiatori di calore, Sant’Ambrogio Servizi Industriali S.r.l., Milano, 2012

15

Metallurgical design of high-toughness, high-strength steels for 300 bar gas cylinders applications

A. Saccocci *, P.E. Di Nunzio *

* RINA Consulting - Centro Sviluppo Materiali S.p.A.

Summary

Current market requires transportable cylinders for industrial gases with improved mechanical properties such as strength, toughness and fatigue resistance. The trend is towards lighter steel cylinders with higher strength levels.

The development of high-strength high-toughness materials for lighter gas cylinders applications requires the metallurgical design of quenched and tempered (Q&T) steels with Ultimate Tensile Strength (UTS) greater than 950 MPa and total elongation greater than 14%.

Considering the limitations on chemical composition due to industrial constraints, two main types of steel grades have been developed and investigated: Cr-Mo and Cr-Mo with V micro-addition.

Literature, commercial and in-house metallurgical models have been used to identify the most suitable chemical compositions for laboratory heat productions following the typical industrial processing route. Both metallographic and mechanical characterizations have been performed and tensile strength of higher than 950 MPa with impact toughness higher than 50 J/cm² have been achieved applying industrial tempering thermal treatments.

1. Introduction

The market requires transportable cylinders for industrial gases with improved mechanical properties such as strength, toughness and fatigue resistance. The trend is towards lighter steel cylinders with higher strength levels. Aim of this work is to design a suitable chemical composition for a steel to be produced as seamless pipe.

The development of high strength high toughness high strength steel grades for lighter gas cylinder applications, such as storage of gaseous hydrogen storage at a pressure up to 300 bar, must be compliant with the EN 1964/2 (ISO 11114-4) standard, i.e. tensile strength >950 MPa, total elongation >14% and average impact toughness at -50°C of 50 J/cm² with a minimum value not below 35 J/cm².

Compared to the manufacturing process from seamless tube, the construction by pressing a flat plate/strip offers some advantages in weight reduction, because the geometrical tolerances are usually better, and the minimum required thickness can be achieved maintaining the maximum thickness to lower levels than those from seamless tubes.

Typical vessels produced from seamless pipes have an outer diameter range from 227 to 229 mm and a thickness in the range 6.8-7.6 mm. Both ranges are representative of the tolerance level of the industrial process. Therefore the mechanical performance of the new steel grade must be able to compensate for this aspect and this makes the target particularly challenging.

2. Metallurgical design

The starting point is the present commercial composition of the steels currently used for 300 bar cylinders, listed in Tab. 1. The components, oil-quenched and tempered for 40 minutes at 600°C, are characterized by a just sufficient toughness level (55 to 60 J/cm² in the average) at -20°C.

16

C Mn Si Ni Cr Mo Al V N

Ref. 1 0.35 0.70 0.20 <0.15 1.0 0.22 0.030 <0.01 <0.011 Ref. 2 0.36 0.75 0.20 <0.15 1.0 0.40 0.030 <0.01 <0.011 Table 1: Chemical composition (mass%) of commercial steels currently used for 300 bar

cylinders.

In order to improve the mechanical characteristics with respect to the commercial grades, besides the chemical composition, also the tempering temperature has been considered as a key parameter in the production process.

The alloy design procedure has been carried out by using literature and in-house empirical models. Variants of the initial compositions have been ranked based on the predicted hardness as a function of the tempering temperature in the range 550-700°C (assuming a fixed time of 40 minutes according to the industrial process). Their critical cooling rate for obtaining a fully austenitic structure have been also evaluated in order to be sure that the oil quenching is able to produce a fully martensitic structure.

In the first optimization step the chemical composition of steels GC1 and GC2 has been formulated (Tab. 2). The former aims at enhancing the toughness performance whereas the latter, alloyed with V, has an improved strength. In steel GC2 the Mo content has been reduced below 0.3 mass% to promote a larger microstructure uniformity and, to avoid the formation of bainite after oil quenching, it has been compensated by adding Cr above 1%.

In these analytical ranges, toughness is not affected.

These compositions have been cast in laboratory (Vacuum Induction Melting, VIM see Fig.1), hot rolled to the final thickness according to a rolling schedule reproducing as close as possible the industrial route, and then subjected to oil quenching and to the corresponding tempering treatments.

.

W1xT1xL (mm) W2xT2xL (mm) Weight (kg) 250x125x250 240x110x250 75

W1

T1

L

W2 T2

7080 mm

Fig. 1: a) Vacuum Induction Melting Facility at CSM; b) Dimension of useful material cast at VIM.

17 A bloom of 75kg of useful material has been cast for each chemical composition, then cut into half and finally hot rolled at reversible duo mill in two phases: roughing mill and finishing mill (Fig2). The temperature evolution during the hot rolling process was measured by manual pyrometer.

t R.R. T Small section 110x115 mm²

(mm) (%) (°C) Large section 120x120 mm²

1 120 - 1250 Lenght 320 mm

2 95 21 - Sprue 100¸110 mm

3 70 26 -

4 49 30 - Weight 35 kg

5 35 29 -

6 25 29 Dimension after roughing

7 20 43 ≥950 21x155x900 mm³

Total R.R. 83

Dimension before finishing

t RR T 21x155x300 mm³

(mm) (%) (°C)

1 14 30 1150 Dimension after finishing

2 10 29 ≥900 11x165x600 mm³

Total R.R. 50

No pass

Dimension before roughing Roughing mill

No pass

Finishing mill

Fig. 2: a) Hot Rolling pilot plant at CSM; b) Dimension of useful material cast at VIM.

C Mn Si Ni Cr Mo Al V N

GC1 0.25 0.80 0.20 0.5 1.20 0.30 0.030 <0.005 0.0050 GC2 0.27 1.00 0.30 0.6 1.50 0.20 0.025 0.060 0.0075 GC3 0.27 1.30 0.30 0.6 0.40 0.60 0.022 <0.005 0.0055 GC4 0.27 1.30 0.30 0.6 0.56 0.40 0.022 0.050 0.0055

Table 2: Chemical composition (mass%) of steels developed for 300 bar cylinders.

The optimum compromise between strength and toughness is GC3.

The microstructure and hardness, in the as-quenched state and after tempering, have been characterized. Both materials produced a fully martensitic structure with a prior austenite grain size of 16 and 12 µm for GC1 and GC2, respectively. An example for the steels GC1 and GC2 is reported in Fig. 3 for tempering at 580 and 700°C.

Then, to check the performance of the model of hardness after quenching and tempering, a comparison between predictions and measurements has been carried out to ascertain its suitability for these grades. The good performance, characterized by a standard deviation of 15 HV30, is illustrated as a scatter plot in Fig. 4. The band at ±30 HV30 is the confidence interval referred to ± 2 standard deviations.

The UTS, estimated from the calculated hardness by multiplying by a factor 3.3, is plotted versus the tempering temperature in Fig. 5. It can be observed that in order to fulfil the minimum requirement for UTS (950 MPa) it is necessary to carry out the tempering treatment at a temperature not higher than about 580°C for steel GC1 and about 650 for steel GC2.

18

a) b)

c) d)

Fig. 3: Microstructure of the Q&T steels in transverse direction: a) GC1 tempered 40 min at 580°C; b) GC1 tempered 40 min at 580°C; c) GC2 tempered 40 min at 580°C; d) GC2

tempered 40 min at 700°C.

19

225 250 275 300 325 350 375 400

225 250 275 300 325 350 375 400 Calculated Hardness (HV30)

Measured Hardness (HV₃₀)

Fig. 4: Performance prediction model for hardness after 40 min tempering in the range 700-550°C starting from a fully martensitic microstructure. The ± 2 standard deviations

confidence band is also reported.

750 800 850 900 950 1000 1050 1100 1150 1200 1250

500 550 600 650 700 750

Calculated UTS (MPa)

Tempering temperature (°C) GC1 GC2 GC3 GC4

Fig. 5: UTS as a function of the tempering temperature calculated from the estimated hardness. The horizontal dashed line is the minimum strength according to the reference

standard for the component.

3. Optimization of the materials performance

Although the strength requirements are fulfilled for both grades, the toughness of these materials can be further improved. The measured impact energy is plotted versus the measured UTS for different tempering treatments in Fig. 6. It can be noticed that the steel GC1 has a good toughness but an insufficient strength, whereas the steel GC2 has a good strength but the impact energy is below the limit of the reference standard.

Therefore, a second optimization step has been undertaken and the alloys GC3 and GC4 have been formulated. It has been decided to increase the Mn content and, concerning the hardenability, to prefer Mo rather than Cr, increasing the former to about 0.5% and

20 decreasing the latter to about 0.5%. Also in this case, the GC4 is a variant alloyed with V to increase the strength. Laboratory casts have been produced and the steels hot rolled and subjected to the Q&T treatments. The as-quenched materials had a fully martensitic structure with a prior austenite grain size of 11 and 12 µm, respectively. An example of the microstructures obtained after tempering is reported in Fig. 7. An example of an additional investigation carried out by SEM on the steels GC3 and GC4 is reported in Fig. 8. Here the two tempering conditions producing similar UTS are compared. The microstructure is typical of a tempered martensite where precipitation of carbides along grain boundaries and within the martensite laths, especially in the V- added variant GC4.

The extremely fine precipitation produced tempering is difficult to be characterized by microanalytical techniques such as SEM-EDS. Therefore an investigation based on commercial thermodynamic softwares (ThermoCalc and JMatPro) has been carried out on all the steels. An example is shown in Fig. 9 where the precipitate fraction as a function of temperature, calculated with the JMatPro software, are shown for all four variants. It has to be noticed that the calculations refer to a full thermodynamic equilibrium and that they do not provide any information on purely kinetic data such as the spatial density and average size of precipitates. Nevertheless they help in the interpretation of the process.

Charpy V-notch Impact Test (transversal specimen, 90° notch)

20 30 40 50 60 70 80 90 100 110 120 130 140

750 800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350 1400 UTS [MPa]

Energy at -50°C[J/cm^2]

A C F2 T3

GC4

GC3 TT=610°C

TT=580°C TT=550°C

TT=610°C TT=640°C

GC1 GC2 GC3 GC4

Fig. 6: Charpy V-notch impact test at -50°C on transversal specimens for all the steels and tempering treatments (TT) versus their respective measured UTS. The dashed box represents the acceptance region according to the reference standard for the component.

21

a) b)

c) d)

Fig. 7: Microstructure of the Q&T steels in transverse direction: a) GC3 tempered 40 min at 580°C; b) GC3 tempered 40 min at 700°C; c) GC4 tempered 40 min at 580°C; d) GC4

tempered 40 min at 700°C.

a) b)

Fig. 8: SEM images of the microstructure of the Q&T steels: a) GC3 tempered 40 min at 580°C; b) GC4 tempered 40 min at 610°C.

It can be observed that the main phases formed are M23C6 and M7C3. The first one forms at higher temperature and, in steels GC3 and GC4, it is the main phase formed during

22 tempering treatments from 700 to about 600. Below this temperature, and in general for all steels, tempering below 600°C produces a mixture of the two phases.

In absence of kinetic effects, it is expected that in the steel GC4 the M7C3 becomes the major precipitate phase below 600°C

a) b)

c) d)

Fig. 9: Thermodynamic evaluation (JMatPro commercial software) of the stability ranges of precipitates formed after tempering in the range from 700 to 500°C:

a) GC1; b) GC2; c) GC3; d) GC4.

To conclude, the comparison of the calculated UTS among all four variants as a function of the tempering temperature has been already reported in Fig. 5. Now it can be noticed that steel GC4 is similar to the corresponding V-alloyed steel GC2 and that the steel GC3 exhibits an intermediate behavior between GC1 and GC2.

The measured UTS and total elongation on the new compositions are shown in Fig. 10 in comparison with their relative lower boundaries according to the reference standard. With a tempering treatment at a temperature less or equal to 600°C, both the steels fulfil the requirements.

Also from the viewpoint of the relationship between impact energy and UTS both materials comply with the standard but, from Fig. 6, it is apparent that GC3 has the best combination of strength and toughness for tempering temperatures of 580 and 550°C.

23

750 800 850 900 950 1000 1050 1100 1150 1200 1250

500 550 600 650 700 750

Measured UTS (MPa)

Tempering temperature (°C) GC3 GC4

10 11 12 13 14 15 16 17 18 19 20

500 550 600 650 700 750

Total elongation (%)

Tempering temperature (°C) GC3

GC4

a) b)

Fig. 10: Measured UTS (a) and total elongation (b) of the steels GC3 and GC4 after quenching and 40 min tempering in the range 550 to 700°C. The horizontal dashed lines are the respective minimum acceptable values according to the reference standard for the

component.

4. Conclusions

The characterization of the chemical compositions selected by exploiting metallurgical models permits to conclude that the optimum combination between strength and toughness, complying with the reference standard for the production of 300 bar gas cylinders, is represented by the steel GC3 quenched and tempered at a temperature in the range from 580 to 550°C for 40 min. The steel processed according to this route is has a UTS in the range between 1100 and 1200 MPa, elongation of 14 to 16% and an average impact energy at -50°C between 50 and 60 J/cm².

This result has been achieved by enhancing the hardenability and the solid solution strengthening of the alloy without exploiting any further hardening due to vanadium additions. This guarantees an excellent homogeneity of the fully martensitic microstructure, as obtained after oil quenching, by avoiding the formation of undesired bainite. In addition, the relative low tempering temperatures promote a fine precipitation of carbides which is fundamental for achieving the target toughness level.

References

[1] “Development of High Strength Steel Plates for Artic Uses Manufactured by Quenching and Tempering Process”, Hisatoshi Tagawa et Al. – Nippon Kokan Technical Report no.44 (1985);

[2] “Développement d’aciers à 100 kg/mm² (980 N/mm²) et haute ténacité en forte épaisseur, pour conduits Forcées” M. SugSuga – Reveue de Métallurgie CIT (Juin 1986);

[3] “Development Of 125ksi Grade Hsla Steel Octg For Mildly Sour Environments”

Masakatsu Ueda et Al. – Corrosion 2005;

[4] UNI EN 1964-2:2001 “Transportable Gas Cylinders - Specification for the Design and Construction of Refillable Transportable Seamless Steel Gas Cylinders of Water Capacities from 0,5 Litre Up to and Including 150 Litres - Part 2: Cylinders Made of Seamless Steel with and Rm Value of 1 100 MPa and Above;

[5] ISO 9809-1:2010 “Gas cylinders -- Refillable seamless steel gas cylinders -- Design, construction and testing -- Part 1: Quenched and tempered steel cylinders with tensile strength less than 1 100 MPa;

[6] “Development and qualification of high-strength high-toughness steel for 300 bar gas cylinder application” A. Saccocci et Al. CSM internal report.

24

Nonlinear Analysis in Pressure Vessel Design Codes Recommendations for Codified Rules Improvements

C. Faidy*

*CF Integrity Consultant

Abstract

During the past 30 years the main rules to design pressure vessels were based on elastic analyses. Many conservatisms associated to these different elastic approaches are discussed in this paper, like: stress criteria linearization for 3-D components, stress classification in nozzle areas, plastic shake down analysis, fatigue analysis, Ke evaluation, and pipe stress criteria for elastic follow-up due to thermal expansion or seismic loads.

This paper will improve existing codified rules in nuclear and non-nuclear Codes that are proposed as alternatives to elastic evaluation for different failure modes and degradation mechanisms: plastic collapse, plastic instability, tri-axial local failure, rupture of cracked component, fatigue and Ke, plastic shakedown. These methods are based on limit loads, monotonic or cyclic elastic-plastic analyses.

Concerned components are mainly vessels and piping systems.

No existing Code is sufficiently detailed to be easily applied; the needs are stress analysis methods through finite elements, material properties including material constitutive equations and criteria associated to each methods and each failure modes.

A first set of recommendation to perform these inelastic analyses will be presented to improve existing codes on an international harmonized way, associated to all material properties and criteria needed to apply these modern methods. An international draft Code Case is in preparation.

1. Introduction

Many Pressure Vessel and Piping Codes, nuclear and non-nuclear, consider “non-linear analyses” at design level as an alternative to linear elastic basic codified approach. One of the major advantages is to limit the discussion on this complex concept of “stress classification in primary versus secondary”. This stress classification can be simple in many cases (shell or beam elements), but can be more complex in a lot of cases (nozzle area or seismic loads…), and in some cases the elastic analysis result can be un- conservative.

Some other applications of the “non-linear analyses” are connected to cyclic loads: plastic shakedown, fatigue analysis, seismic loads or other dynamic cyclic loads.

This paper, developed under EPERC potential Task Group on "Nonlinear Design Rules", only considers components outside of creep regime. It does not cover in this edition buckling and very high seismic loads.

The codes consider in the initial comparison report [Ref. 2 to 10] are:

- Non-nuclear Codes: ASME Boiler & Pressure Vessel Code Section VIII and EN 13445 European Standard,

- Nuclear Codes with English version available: RCCM, RCC-MRx, ASME Boiler &

Pressure Vessel Code Section III and KTA German Code,

- Nuclear Codes with available information in literature (but no English Code Edition available): JSME Pressure Vessel Code, Russian Nuclear Codes and R5 Rule (not strictly a Code)

Five major failure modes are considered:

25 - excessive deformation or plastic collapse

- plastic instability - local failure - rupture buckling (later)

Two major degradation mechanisms are considered:

- Fatigue (Ke, K…) and cycle by cycle evaluation - Plastic shakedown

Other applications are proposed in some Codes for Stress classification and elastic follow- up in particular for piping systems. Three types of loads are considered:

- monotonic loads - cyclic loads

- seismic dynamic-cyclic loads (generally associated to specific rules for piping) Final report organization is (Figure 1):

- Part 1: Code Comparison and Open points [11]

- Part 2: Recommended practices - Part 3: Benchmarks [12]

This paper reviews the status of the "Recommended Practices" document that makes a set of proposals for a harmonized Code Case that will be managed by each Code Organization. Some examples of "Recommended Practices" are presented in this paper, the others that needs complex cyclic material constitutive equations are on the finalization process.

2. General open points in the different codes

The Code comparison done in Part 1 is associated to a set of "Open points" to be solved in the "Recommended Practices" report:

- Tresca versus Von Mises plasticity criteria and need of correction factor on limit load, - No guidance for the use of Finite Element Analysis (FEA) for limit load evaluation, - Limits on the use of limit analysis for level D criteria (possible large displacement)

and flow stress value,

- Detailed recommendations for analysis, as:

o Temperature to use for each step of these analyses

o Tolerances and precise geometry to perform inelastic analysis

o More general FEA recommendations (mesh refinement, small versus large displacement, convergence criteria…)

o Cyclic elastic-plastic analyses for fatigue and shakedown

o Strain criteria for elastic-plastic analysis; the criteria have to be in the Code and can be modified case by case by the user’s design specification.

o Local failure requirements: in level D or through elastic-plastic analysis

o All the material properties for all these analysis (engineering/ true and cyclic stress-strain curves, material constitutive equations…)

o Thermal expansion load classification in piping systems

o Requirement on satisfaction of the wall thickness requirements o Different Sy / Su limits

o K, Ke and Kn factors: optimization and associated justification of simplified elastic-plastic analysis

o Cyclic elastic-plastic analyses for fatigue and shakedown needs more detailed procedure

26

3. Conclusion on code comparisons

All codes define limitations to justify component failure margins from the application of operation mechanical or thermal loads.

The failure mechanisms which are accounted for are the following:

- excessive deformation (plastic collapse) - plastic instability

- local failure (decohesion) - rupture

The degradation mechanisms considered, generally associated to cyclic loading are:

- progressive deformation induced by repeated loads (ratchetting/shakedown) - fatigue

Some Codes consider the damage mechanism connected to stress triaxiality: local failure by decohesion, clarification of background and needs of alternative nonlinear rules is required.

All the Codes considered in this report have to be strongly improved in order to assure clear and precise use of different types of non-linear analysis, be more efficient for some particular case and take benefit for less conservative rules than elastic analysis.

The different non-linear analyses are:

- limit analysis associated to elastic perfectly plastic material; the corresponding criterion are based on load comparison,

- monotonic elastic-plastic analysis associated to material stress-strain curve; the corresponding criterion are based on a maximum strain level (sufficiently low compared to material maximum elongation),

- cyclic elastic plastic analysis associated to material cyclic stress-strain curve for fatigue plasticity correction factor or more sophisticated material constitutive equation for shakedown and ratchetting analysis.

Some other applications of non-linear analysis are associated to "stress classification"

improvements and alternative rules, like:

- pipe / nozzle neck area,

- piping systems thermal expansion stress classification (elastic follow-up), - piping systems under seismic loads, in particular for high level seismic event.

The review of all the existing Codes and existing non-linear analysis design rules [11] are key issue to start, develop the “best practice document” (Report Part 2), associated to detailed procedure proposals and validation- background at each level of the procedures.

4. Preliminay proposals for the recommended practices document

The scope of the document will cover:

- Components: Vessel, piping; nuclear class 1 and 2, and non-nuclear - Failure Modes: Plastic collapse, Instability, Local Failure

- Degradation mechanism: fatigue, plastic shakedown

- Analysis Methods: Non-linear analysis (limit analysis and elastic plastic) - Loads: monotonic, cyclic, dynamic

- Type of loads: mechanical (dead weight, pressure…), thermal (thermal expansion, thermal gradient transients), seismic…

- General Quality Management: Requirements for Computer Codes and Analysts After an overview of the different methods with detailed definition and validity limits, the report will propose recommendations to perform these type of analyses: limit load analysis, monotonic and cyclic elastic-plastic analyses.

27 The next part of the report will develop methods, criteria and all the material properties needed to perform the different analysis for: plastic collapse, plastic instability, local failure, fatigue and plastic shakedown.

For plastic collapse, two methods will be proposed: plastic limit load with the material yield strength or an elastic-plastic analysis with "double–slope" criteria.

For plastic instability, one method using elastic-plastic analysis with strain criteria and another one (not completely theoretically valid) based on limit load with a flow stress equal to half the yield strength plus maximum strength of the material are proposed. Some particular attention will have to consider possible large displacement of some structures (as piping systems).

For rupture: reference stress evaluation (limit load of cracked components), and more innovative models as local approach of rupture (more details later) are discussed.

For fatigue two ways will be proposed: one by simplified analysis using Ke evaluation with cyclic stress-strain curve and kinematic hardening and the other one directly connected with the plastic shakedown analysis by direct cycle by cycle analysis and dedicated plasticity laws for the material (evolutive mixed hardening laws…).

For plastic shakedown a direct elastic perfectly plastic analysis could be a first step approach.

5. Definitions

5.1 Elastic-Plastic Analysis

Elastic-Plastic analysis is that method which computes the structural behavior (stress, strain, displacement…) under given loads considering the plasticity characteristics of the materials, including strain hardening (or strain softening) and the stress redistribution occurring in the structure.

5.2 Limit Load Analysis

Limit analysis is a special case of plastic analysis in which the material is assumed to be ideally plastic (non-strain-hardening). In limit analysis, the equilibrium and flow characteristics at the limit state are used to calculate the collapse load. This method requires "radial" (or "proportional") applied loads.

The two bounding methods which are used in limit analysis are:

- the "lower bound approach", which is associated with a "statically admissible" stress field,

- and the "upper bound approach", which is associated with a "kinematically admissible" velocity field.

The consequences are:

- the Finite Element Method computer codes use a "displacement" model, a

"kinematically admissible" velocity field, and leads to an "upper bound" limit load results

- only few computer codes use a "stress model", a "statically admissible" stress field, and can lead to a "lower bound" limit load results that is better in term of codified rules

- all the background theory requires:

o a radial (or proportional) load, incremental load by multiplication by  from the elastic regime up to the limit load

o the plasticity criteria and the displacement continuity have to be fulfilled

o mainly used for plastic collapse on the initial geometry with a flow stress equal to the yield stress of the material

28 o for higher load analyses, as "plastic instability" based on an intermediate flow stress between the yield stress and the maximum strength of the material, the possible interaction between "large plastic strain" and "large displacement"

(significant change of the geometry compared with the initial geometry of the component) has to be considered by the user

5.3 Collapse Load

Definition: no general excessive deformation (not over the yield strain limit used)

An elastic-plastic analysis may be used to determine the collapse load for a given combination of loads on a given structure. The following criterion for determination of the collapse load shall be used (Figure 2):

- a load–deflection or load–strain curve is plotted with load as the ordinate and deflection or strain as the abscissa. The angle that the “linear part of the load–

deflection or load–strain curve” makes with the ordinate is called .

- a second straight line, hereafter called the “collapse limit line”, is drawn through the origin so that it makes an angle  = tan^-1(2 tan ) with the ordinate.

The collapse load is the load at the intersection of the load–deflection or load–strain curve and the collapse limit line. If this method is used, particular care should be given to ensure that the strains or deflections that are used are indicative of the load carrying capacity of the structure.

An alternative method to determine the collapse load can be used: limit analysis with elastic perfectly plastic stress-strain curve using the material yield strength.

5.4 Plastic Instability Load

The plastic instability load for areas under predominantly tensile (or compressive) loading is defined as that load at which unbounded plastic deformation can occur without an increase in load. At the plastic tensile instability load, the true stress in the material increases faster than strain hardening can accommodate (Figure 3).

Collapse Load: The methods of limit analysis are used to compute the maximum load that a structure assumed to be made of ideally plastic material can carry. At this load, which is termed the “collapse load”, the deformations of the structure increase without bound.

Collapse Load — Lower Bound: If, for a given load, any system of stresses can be found which everywhere satisfies “equilibrium”, and “nowhere exceeds the material yield strength”, the load is “at or below the collapse load”. This is the lower bound theorem of limit analysis which permits calculations of a lower bound to the collapse load.

Plastic Hinge: A “plastic hinge” is an idealized concept used in Limit Analysis.

In a beam or a frame, a plastic hinge is formed at the point where the moment, shear, and axial force lie on the yield interaction surface.

In plates and shells, a plastic hinge is formed where the generalized stresses lie on the yield surface.

5.5 Local failure (decohesion)

In all location in the vessels, including nozzle areas, in level A-B-C the stress triaxiality has to be limited.

5.6 Rupture

For some key pressure equipment, the resistance to hypothetical large crack has to be considered at the design level or end of fabrication crack size has to be evaluated