The Experimental Work in Support to ATLAS+ Project

Within the framework of European project MULTIMETAL (Structural performance of multi-metal component), several fracture tests on different types of multi-material specimens have been performed. Present fracture toughness standard methods, e.g. ASTM E 1820 are not directly intended for Dissimilar Metal Weld (DMW). Therefore further investigations are needed in order to define the best practice in fracture mechanical tests and their analysis for DMWs. Specimens are taken from welded plates: a narrow gap Inconel DMW junction between ferritic and austenitic stainless steels, designed and delivered by AREVA France. The aim of this work is to provide guidelines for the determination of DMW fracture properties. For that purpose, fracture specimen needs to be modelled by FE. The first task, which is the purpose of that paper, is the determination of the mechanical properties in terms of stress-strain curve of all DMW constitutive materials: austenitic stainless steel, ferritic steel, heat affected zone of the ferritic steel zone and Nickel alloy zone properties. Conventional techniques for tensile test are not able to provide the tensile curve of the different materials constituting a weld joint. Image correlation techniques are well suitable but imply too long and difficult work for the images analysis. Therefore CEA has developed an intermediate solution based on laser sensors which provides a complete profile of the specimen during the tensile test. Using Bridgman equations, the stress and strain can be deduced from the measurement of the shape of the specimen (reduction of section but not only…). This innovative device has been used with new developments using local Bridgman equations in the post-processing of measurements. This allows to access to the material behaviour of several materials with only one specimen. Numerical interpretation using FE methods is presented and confirms the material behaviour determined from the experimental work using Bridgman equations assumptions. Finally, this combined experimental and numerical work has provided material data relative to all constitutive materials of the DMW junction. A hardened area in stainless steel material due to the welding process has been pointed out, and the heat affected zones of the ferritic material have been characterized in terms of stress-strain curves. The next stage of the project is to carry out tests on fracture specimens and to model these multi-materials specimens by FE. The gradient of elasto-plastic properties is now available for this next step.

Typically in flaw evaluation procedures, idealized crack shapes are assumed for both subcritical and critical crack analyses. Past NRC-sponsored research have developed estimation schemes for predicting the load-carrying capacity of idealized cracks in nuclear grade piping and similar metal welds at the operating conditions of nuclear power reactors. However, recent analyses have shown that growth of primary water stress corrosion cracks (PWSCC) in dissimilar metal (DM) welds is not ideal; in fact, very unusual complex crack shapes may form, i.e., a very long surface crack that has a finite length through-wall crack in the same plane. Even though some experimental data on base metals exists to demonstrate that complex shaped cracks in high toughness materials fail under limit load conditions, other experiments demonstrate that the tearing resistance is significantly reduced. At this point, no experimental data exists for complex cracks in DM welds. In addition, it is unclear whether the idealized estimation schemes developed can be used to predict the load-carrying capacity of these complex-shaped cracks, even though they have been used in past analyses by the nuclear industry. Finally, it is unclear what material strength data should be used to assess the stability of a crack in a DM weld. The NRC Office of Nuclear Regulatory Research, with their contractor Battelle Memorial Institute, has concluded an experimental program to confirm the stability behavior of complex shaped circumferential cracks in DM welds. A combination of full-scale pipe experiments and a variety of laboratory experiments were conducted. A description of the pipe test experimental results is given in a companion paper. This paper describes the ongoing analyses of those results, and the prediction of the load-carrying capacity of the circumferential cracked pipe using a variety of J-estimation scheme procedures. Discussions include the effects of constraint, appropriate base metal material properties, effects of crack location relative to the dissimilar base metals, and the limitations of the currently available J-estimation scheme procedures. This paper concludes with plans for further development of J-estimation scheme procedures for circumferential complex cracks in DM welds.

Multilevel Methods for the Simulation of Turbulence: A Simple Model

A fitness-for-service (FFS) evaluation was performed on a cogeneration power plant high pressure steam line. The piping system included a Gr 91 to Gr 22 dissimilar metal weld (DMW) between an SA-335 Gr P91 pipe and an SA-335 Gr P22 reducer. This unit had accumulated more than 100,000 service hours. The FFS project included as-built and as-found piping stress analyses, redistribution of creep stresses, weld residual stress estimates, review of the nondestructive examination (NDE) results, evaluation of the plant information data, and a fracture mechanics evaluation. Due to tapering of the Gr P22 pipe to match the wall thickness of the Gr P91 pipe at the DMW interface, the Gr P22 pipe wall thickness was not in compliance with the ASME B31.1-2000 Code minimum wall thickness (MWT). In addition, an inspection of the single groove girth DMW revealed ultrasonic indications along the fusion face of the weld. The lack of side wall fusion (LSWF) can be considered as a planar and crack-like indication, so a fracture mechanics evaluation was performed. The piping stress analysis revealed that the DMW was subject to low axial stresses. A creep life fraction evaluation was performed for the material that does not have the LSWF indication, considering conservative creep rupture data, weld performance degradation, redistributed creep stresses, and multiaxial stresses in the DMW. Due to the low stresses, the life fraction analysis estimated more than 50 years of remaining creep life. Failure Assessment Diagrams (FADs) were developed to evaluate the component integrity for various combinations of crack depths and lengths. The reported indications should not compromise the integrity of the component subject to the typical plant operation during the next 12 years and the component is fit for continued operation. The crack propagation evaluation determined that the DMW has a predicted minimum safe operating life of at least 12 years if operated as in the past. Because the wall thickness of the tapered reducer is below the ASME B31.1 Code allowable MWT, it was recommended that the DMW be examined during a scheduled outage within 5 years and the component be replaced with Gr P91 material during a major outage within 8 years.

This paper presents a preliminary study using probabilistic fracture mechanics codes (PRO-LOCA 2009 and WinPRAISE 2007) to estimate the rupture probability of Dissimilar Metal Weld (DMW) in CANDU^[!R] small diameter feeder piping subject to potential Primary Water Stress Corrosion Cracking (PWSCC). In spite of the significant differences in the deterministic fracture mechanics models and algorithms used for speeding up Monte Carlo calculations, both PFM codes consistently conclude the following observations: ・The rupture probability of a single DMW under normal operating conditions is extremely small. ・Timely response to leakage can substantially reduce the rupture probability. ・In-service inspection (every three years) has insignificant effect on the rupture probability. ・Residual stress (231 MPa vs 400 MPa) has insignificant effect on the rupture probability. ・The leak probability is several orders of magnitude larger than the rupture probability. The mode of failure of DMW is leak-before-break.

The paper presents estimated thermal expansion coefficients of 316SS and 508LAS base, 316SS filler or similar metal weld (SMW), and 316SS-508LAS dissimilar metal weld (DMW) filler and butter weld metals. These base and weld metals are typically used in nuclear reactor pressure boundary components. Accurate estimation of the expansion coefficients of these metals is essential for accurate estimation of thermal-mechanical stress in reactor pressure boundary components. In this paper we present the expansion coefficients of 316SS and 508LAS base, 316SS-SMW filler, and 316SS-508LAS DMW filler and butter weld metals. The coefficients were estimated based on our own experimental data. The corresponding expansion coefficient results and the FE validation results are presented in this paper. We anticipate that these types of results can be used as guidelines for choosing appropriate expansion coefficients for thermal-mechanical stress analysis of safety critical nuclear reactor components.

Fracture mechanics safety analysis of components based on fracture mechanics characteristics combined with multiaxiality of the stress state

Incremental unknowns, multilevel methods and the numerical simulation of turbulence

A program of experimentation and analysis aimed at determining fracture instability in circumferentially cracked Type 304 stainless steel pipes in seismic or water-hammer loadings is described. Experimental work on center-cracked tension panels revealed that dynamic loading does not affect the net-section stress criterion evolved previously for Type 304 stainless steel. Full-scale tests on 100-mm-diameter (4 in.) pipes subjected to a dynamic load nevertheless indicate that a margin of safety exists beyond that predicted by the application of the net-section stress criterion. It is concluded that the finite duration of a dynamic loading together with the system compliance which allows stable crack growth beyond maximum load is primarily responsible. A J-based plastic fracture mechanics assessment based upon rate-dependent mechanical and fracture properties of the material was made and found to be consistent with this hypothesis.

Review of empirical and semi-empirical Y factor solutions for cracked welded tubular joints

Crack increment is a phenomenon in fracture mechanics. It is occurred due to the stress concentration at the imperfection material. Thus, it leads a crack to growth. Then, the crack will reach to a critical crack length before catastrophic failure could occur. Before the catastrophic failure occur, the cracked structure can be fully utilised until the crack reach to the critical crack length. Thus, it is crucial to investigate the behaviour of the crack increment in fracture mechanics. The main purpose of this paper is to model the crack increment in fracture mechanics analysis via Kolmogorov-Smirnov test. The modelling requires a collection of crack data through experimental work. Then, the data was evaluated based on Kolmogorov-Smirnov test. The results show that the crack increment can be modelled as a Gaussian distribution.

The heterogeneity of welded joints' microstructure affects their mechanical properties, which can vary significantly in relation to specific weld zones. Given the dimensional limitations of the available test volumes of such material zones, the determination of mechanical properties presents a certain challenge. The paper investigates X welded joint of S690QL1 grade high strength steel (HSS), welded with slightly overmatching filler metal. The experimental work is focused on tensile testing to obtain stress-strain properties, as well as fracture mechanics testing. Considering the aforementioned limitations of the material test volume, tensile testing is carried out with mini tensile specimens (MTS), determining stress-strain curves for each characteristic weld zone. Fracture mechanical testing is carried out to determine the fracture toughness using the characteristic parameters. The experimental investigation is carried out using the single edge notch bend (SENB) specimens located in several characteristic welded joint zones: base metal (BM), heat affected zone (HAZ), and weld metal (WM). Fractographic analysis provides deeper insight into crack behavior in relation to specific weld zones. The numerical simulations are carried out in order to describe the fracture behavior of SENB specimens. Damage initiation and evolution is simulated using the ductile damage material behavior. This paper demonstrates the possibility of experimental and numerical determination of fracture mechanics behavior of characteristic heterogeneous welded joint zones and their influence on crack path growth.

Application of fracture mechanics methods in safety analysis of piping components in subcreep and creep behavior

Characterization of the fracture resistance of weld joints, and in particular dissimilar metal welds (DMW), is a huge and difficult work where no standard currently exists. As a consequence, characterizations of materials have to be done prior to the fracture tests in order to consider the mechanical aspects of the material mismatch via F.E. analysis in the fracture toughness determination. However, performing these characterization tests may imply using a lot of material, which is sometimes not available, and can be expensive. Most studies are usually realized on reduced sized mock-ups. The considered experimental mock-ups are pipes composed of a ferritic pipe welded to an austenitic one. The weld joints are made of Ni base alloys. A few small tensile samples have been extracted so that each material, even the heat affected zone (HAZ), can be characterized at low temperature. A fast method using laser sensors and a specific specimen shape has been developed and is used to identify these materials strain-stress curves at −125°C. Afterwards, these data have been used to simulate multi-material compact specimen tensile tests and single edge notch tension specimen, representing a conventional defect in the HAZ. Also, these two kinds of specimen have been extracted and the tests performed in order to compare the experimental results to the F.E. analysis. This paper presents the experimental work, the related specific devices, the F.E. analyses and the experimental analysis.

In the nuclear industry narrow gap welding techniques are used to perform junctions between ferritic low alloy steel heavy section components and austenitic stainless steel piping systems. The residual stresses in Dissimilar Metals Welds (DMW) may influence the lifetime and functionality of the welded components. In Pressurized Water Reactor (PWR) piping systems, weld residual stresses in particular increased the susceptibility to primary water stress corrosion cracking (PWSCC) in the past. It is therefore necessary to develop and validate methods for a reliable residual stress and distortion prediction. Numerical welding simulations for predicting residual stresses are commonly used in nuclear industry and their development is progressing fast during the recent years. As part of the European project STYLE — Structural Integrity for Lifetime Management — a case study was launched to assess the capability of such simulations. The mock-up in this case study is provided by AREVA NP SAS. It is a pipe with a narrow gap dissimilar metal weld. The pipe thickness is about 40 mm and the outer diameter is 352 mm after final machining. In this assembly a 316L austenitic pipe is welded to an A508 Class 3 ferritic pipe by means of Alloy 52 Gas Tungsten Arc (GTA) narrow gap weld which is representative for PWR primary circuit piping. This mock-up is in the scope of a continuation of the ADIMEW – Assessment of Aged Piping Dissimilar Metal Weld Integrity - project and deals with the improvement of the assessment for DMW and Leak-before-break (LBB) procedures. The fracture test on this mock-up is planed to be performed at 300 °C with an initial through-wall defect. Apart from the LBB demonstration this mock-up is also dedicated for the validation of the applied fracture mechanics approach, extension of material data basis and validation of the weld simulation procedures applied within AREVA. This paper presents the results of the finite element residual stress analysis related to this STYLE narrow gap weld case study. The two finite element codes ABAQUS and SYSWELD were used to predict the weld residual stresses and the shrinkage in axial direction. The major difference between the here presented methods is that SYSWELD accounts for phase transformation and the method used with ABAQUS does not. The results are compared between each other and with data obtained by deep-hole drilling techniques (DHD) at several locations.