Fracture toughness assessment of in-service aged primary circuit elbows using mini-CT specimens taken from outer skin

doi:10.1016/S0029-5493(97)00363-4

Nuclear Engineering and Design

Volume 184, Issue 1, 1 August 1998, Pages 3-11

https://doi.org/10.1016/S0029-5493(97)00363-4 Get rights and content

Abstract

Type CF8M cast duplex stainless steels used in the primary circuit elbows of pressurized water reactors are subject to thermal aging embrittlement at their service temperature, at ∼300°C. This phenomenon affects their fracture toughness properties. In order to assess the residual fracture toughness of these elbows, estimations are made through predictive formulae based on chemical composition and aging conditions, which provide safe values. However, in the case of the most sensitive materials, it is important to obtain more accurate estimations. A new method of determination was thus considered, based on the testing of mini-CT specimens taken from the skin of in-service elbows. The feasibility of using mini-CT specimens to evaluate the tearing resistance of cast duplex stainless steels seems at first sight difficult, in particular because of the very coarse metallurgical structure of these steels: will small specimens be representative of larger volumes (mainly regular 1T–CT specimens) and will they not induce too much scatter? In order to answer such questions, an experimental validation program has been undertaken: the results show that the method is relevant and lead to propose guidelines which aim at optimizing the experimental data analysis. Then the method is applied to an in-service elbow: the results obtained are found to be in good agreement with the toughness estimations given by our predictive formulae. This subsequently contributes to the validation of the general methodology used for the justification of French primary circuit elbows.

Introduction

Type CF8M cast duplex stainless steels used in the primary circuit elbows of pressurized water reactors are known to undergo thermal aging embrittlement at their service temperature, at ∼300°C. This phenomenon affects their mechanical properties, mainly their fracture toughness (tearing resistance) (Bethmont et al., 1991, Massoud et al., 1991), which needs to be evaluated for mechanical analyses.

Fracture toughness estimation methods through predictive formulae using chemical composition and aging conditions are available (Slama et al., 1983), which provide safe values. However, in the case of the most embrittled elbows, more accurate evaluations are also needed. A new method was therefore considered: it consists of taking relatively thin slices of material from elbow surfaces and performing tearing resistance tests on very small fracture mechanics specimens (0.4T–CT with thickness reduced to 5 mm), called mini-CT.

Given the coarse macrostructure of the cast duplex stainless steels and the level of scatter usually encountered on tearing resistance curves using 1T–CT specimens, the use of mini-CT specimens is not obviously relevant. A validation program was thus undertaken before starting taking samples from in-service elbows.

This paper deals with the development and application of mini-CT specimens testing. It is divided into three parts: the first one presents the experimental procedures, the second shows the results of the validation program and the third is devoted to the application to an in-service elbow.

Section snippets

Experimental procedures

All tearing resistance tests were carried out according to the GFR procedure (French Fracture Group, 1990). Except for a few differences, this procedure is equivalent to the ASTM E813-89 standard.

Validation program

In order to check the relevance of tearing resistance tests using mini-CT specimens, a validation program was undertaken. Our major concern was the level of scatter of the J−Δa results, which we feared might be enhanced by reducing the affected volume of material: there are only one or two primary ferritic grains ahead of the crack front of mini-CT specimens, whereas they are five times as numerous for 1T–CT specimens (Jayet-Gendrot et al., 1994).

The validation program consists of testing

Application to an in-service elbow

Three samples of approximate dimensions 55×54×6 mm were taken from the outer skin of a hot leg 50° elbow in Gravelines 4 reactor. The choice of sampling areas was guided by industrial considerations (sufficient thickness, accessibility, absence of repair and of radiographic indication, presence of full lagging). This elbow had remained in service for 86 898 h at the temperature of 323°C. The ferrite contents measured on the samples range from 33 to 35%. The chemical composition measured on one

Conclusion

An experimental program was carried out to study the use of small 0.4T–CT specimens with reduced thickness to estimate the fracture toughness of embrittled cast materials. The results showed that such mini-CT specimens are relevant and guidelines for further practice are proposed. The main outcomes are the following:

•
the J−Δa results obtained with mini-CT specimens are correctly centered around those obtained with regular 1T–CT specimens. Scatter is of the same order.
•
the results corresponding to

Acknowledgements

We would like to thank M. Naumann for her work on specimen observations during her internship at EDF.

References (7)

Bethmont, M., Meyzaud, Y., Soulat, P., 1991. Properties of Cast Austenitic Materials for Light Water Reactors. Proc....
French Fracture Group, 1990. Test Recommendations for Measurement of the Ductile Tearing Resistance of Metallic...
Jayet-Gendrot, S., Ould, P., Balladon, P., 1994. Effect of Fabrication and Test Parameters on the Fracture Toughness of...

There are more references available in the full text version of this article.

Cited by (16)

Microstructure evolution and mechanical behavior of a lean duplex stainless steel aged at 475<sup>o</sup>C
2021, Materials Science and Engineering: A
Citation Excerpt :
Over the past five decades, a significant amount of research has been published on 475 °C embrittlement of DSSs, showing that the plasticity of spinodal decomposed ferrite is reduced because of a reduction in the number of slip systems and that deformation by twinning may also take place in addition to deformation by dislocation glide. These microstructural changes can alter, among several properties, tensile strength and hardness [7–18], fracture toughness [19–38] and fatigue behavior of the alloy [39–71]. However, some open questions still justify the interest on microstructural evolution and changes in the mechanical properties of DSSs before and after thermal aging.
In this research, the effects of isothermal treatments at 475 °C on the microstructure and mechanical properties of a UNS S32304 lean duplex stainless steel were investigated. Samples of the alloy under as received and aged conditions were analyzed by atomic force microscopy and magnetic force microscopy in order to evaluate the ability of these techniques to identify important aspects that characterize microstructural changes caused by aging. The mechanical behavior was evaluated by tensile tests, hardness tests, Charpy impact tests, crack extension resistance tests, force controlled axial fatigue tests and fatigue crack growth rate tests. The fracture surfaces of all tested specimens were analyzed by scanning electron microscopy. The results indicate that the α-phase spinodal decomposition occurred, with a tendency to saturation for a not too long aging time. The analysis by atomic/magnetic force microscopy allowed to identify characteristics regarding the preferential phase dissolution of the α phase that distinguish the alloy in the as received condition from the aged alloy. As a consequence of the formation of the α’ phase, the absorbed impact energy and the fracture toughness of the alloy decreased substantially. On the other hand, the tensile strength and the fatigue resistance increased significantly.
Prediction of J-R curves and thermoelectric power evolution of cast austenitic stainless steels after very long-term aging (200,000 h) at temperatures below 350 °C
2020, Journal of Nuclear Materials
Citation Excerpt :
This set of data is very important in our process as it was used to validate the prediction model. It counts more than ten characterizations of removed components (hot or cold elbows), to which must be added a dozen of characterizations from specimens taken from the outer skin of primary circuit elbows [2]. Because the relative contributions of the aging mechanisms (spinodal decomposition, G-phase precipitation …) to the evolution of Charpy-impact energies and TEP values may be different, the apparent activation energies determined for each property are not necessarily identical.
Cast austenitic stainless steels (CASS) are materials used to fabricate many important safety-related components in the primary circuits of light water reactors since the early 1970’s. The primary circuit, which transports heated water by nuclear reaction to steam generators, is subjected to in-service temperatures between 285 °C and 325 °C. Under these conditions, CASS undergo thermal aging which may significantly affect their mechanical properties and more especially their fracture toughness. From a metallurgical point of view, the changes in mechanical properties are attributable to several solid-state phase transformation processes including the spinodal decomposition of the ferritic phase and the precipitation of G phase. The kinetics of these phase transformations depends primarily on the time and temperature of aging, secondly on the chemical composition of the heat (chromium, molybdenum, silicon and nickel contents) and thirdly on the heat treatments performed during component manufacturing. The prediction of the long-term behavior of CASS is an important industrial issue for nuclear power plant operators. In the early 1980’s, Electricité De France (EDF) engaged an unparalleled laboratory aging program that is still in progress. Presently, some materials have been aged for more than 200,000 h at low temperatures. The results of this program allow the development of prediction models for Charpy-impact energies, $J - R$ curves and thermoelectric power values much more precise than those proposed so far, based on aging at 400 °C and rarely exceeding 30,000 h. These prediction models are described in this paper.
A review of irradiation effects on LWR core internal materials - Neutron embrittlement
2011, Journal of Nuclear Materials
Citation Excerpt :
For these specimens, J values above 150 kJ/m2 and crack extensions beyond about 1.2 mm are above the validity limits based on ASTM Specification E 1820-06. However, comparison of fracture toughness data obtained on 1-T CT and small specimens indicates that small specimens yield equivalent J–R curve data at least for materials with JIc values up to about 200 kJ/m2 [41]. This relationship has been used to determine JIc in most investigations on neutron embrittlement [14,43].
Austenitic stainless steels (SSs) are used extensively as structural alloys in the internal components of light water reactor (LWR) pressure vessels because of their relatively high strength, ductility, and fracture toughness. However, exposure to neutron irradiation for extended periods not only changes the microstructure and microchemistry of these steels, but also degrades their fracture properties. The existing data on irradiated austenitic SSs are reviewed to determine the effects of key parameters such as material type and condition and irradiation temperature, dose, and dose rate on neutron embrittlement. Differences in the radiation-induced degradation of fracture properties between LWR and fast-reactor irradiations are also discussed. The results are used to (a) define a threshold fluence above which irradiation effects on fracture toughness of the material are significant, (b) evaluate the potential of neutron embrittlement under LWR operating conditions, and (c) assess the potential effects of voids on fracture toughness.
Effect of 475 °C embrittlement on the mechanical properties of duplex stainless steel
2009, Materials Science and Engineering: A
The binary iron–chromium alloy embrittles in the temperature range of 280–500 °C limiting its applications to temperatures below 280 °C. The embrittlement is caused by the decomposition of the alloy to chromium-rich phase, α′ and iron-rich phase, α. This phenomenon is termed 475 °C embrittlement as the rate of embrittlement is highest at 475 °C. Primarily the investigations on 475 °C embrittlement were confined to binary iron–chromium alloys and ferritic stainless steels. Duplex stainless steel grades contain varying proportions of ferrite and austenite in the microstructure and the ferritic phase is highly alloyed. Moreover, this grade of steel has several variants depending on the alloy composition and processing route. This modifies the precipitation behaviour and the resulting change in mechanical properties in duplex stainless steels when embrittled at 475 °C as compared to binary iron chromium systems. The precipitation behaviour of duplex stainless steel at 475 °C and the effect on tensile, fracture and fatigue behaviour are reviewed in this article.
The effect of ageing temperature and time on the mechanical properties of Fe-NiCrMo alloys with different contents of δ ferrite
2008, Nuclear Engineering and Design
Laboratory cast alloys with 2–27% of δ ferrite were aged for up to 17,520 h in the temperature range 290–350 °C. Tensile and Charpy tests were performed at 22 and 290 °C on specimens aged for different times, and the microhardnesses of both constituents of the microstructure were determined for the alloy with 27% of δ ferrite. The effects of the content of δ ferrite, the ageing and testing temperature, and the ageing time on mechanical properties and notch toughness are presented and discussed.
Damage and dynamic strain aging in a thermal aged cast duplex stainless steel
2004, Journal of Nuclear Materials
An experimental study was carried out at 320 °C in order to determine the damage kinetics of a thermal aged duplex stainless steel. Compression and tensile tests were performed at 320 °C and at 20 °C for comparison. Typical steps of damage for this material were found at room temperature. At 320 °C, no bulk damage was observed, but strain and damage were localized. The experimental mechanical results revealed the existence of dynamic strain aging. The absence of bulk damage is linked to this phenomenon: as soon as damage appears in ferrite, there is localization and fracture follows.

View all citing articles on Scopus

View full text