Liquid metal embrittlement of the martensitic steel 91: influence of the chemical composition of the liquid metal.: Experiments and electronic structure calculations
Introduction
The occurrence of liquid lead embrittlement of a martensitic steel Z10 CD Nb V 9-1 (AFNOR) steel (designated hereafter steel or grade 91) at temperatures close to the lead melting temperature has been recently reported [1], [2]. To prompt the embrittlement, the adopted strategy consisted in increasing the material strength by using a well adapted thermal treatment and, combined to this, it was also necessary to machine a notch on the initially smooth cylindrical tensile specimens. Our experimental results clearly showed that the embrittlement by liquid lead disappears if the testing temperature is raised above 450 °C [2].
In this work we are dealing with results concerning the embrittlement of the grade 91 (submitted to a hardening heat treatment) by different liquid metals (the eutectic Pb–Bi, Sn and Hg). The corresponding binary phase diagrams Fe–liquid metal are different: the Fe–Pb–Bi and Fe–Hg systems on one side are nonmiscible while on the other side the Fe–Sn system has intermetallics like FeSn and FeSn2. As will be shown, however, the mechanical response of the grade 91 in contact with the liquid metals cannot be deduced from these simple thermodynamic considerations sinceas far as the mechanical behaviour is concerned, the Fe–Pb–Bi system is closer to the Fe–Sn system than to the Fe–Hg one.
Given the experimental conditions adopted in this work, the observed embrittlement is very likely to be a direct consequence of the reduction of the steel surface energy due to the adsorption of liquid metal atoms as pointed out by Rehbinder and co-workers for instance [3]. The reduction of the surface energy by adsorption of liquid metal atoms can in principle be evaluated from atomic scale simulations relying on accurate models for the interaction potential energy between the different kinds of atoms. Here, we have performed electronic structure (ab initio) calculations to address the chemical interactions between the (1 0 0), (1 1 0) and (1 1 1) iron surfaces and Hg, Pb, Bi and Sn atoms. It will be shown that the calculation of the adsorption energies can give useful trends towards an LME prediction.
Section snippets
Base material
The base material supplied by Creusot Loire Industries is the Z10 CD Nb V 9-1 (AFNOR) steel designated steel or grade 91. Its chemical composition is given in Table 1. The as received material has a tempered martensitic microstructure resulting from the standard heat treatment consisting in heating 1 h at 1050 °C, air quenching, heating 1 h at 750 °C and then air cooling. Due to its low carbon content, its crystallographic structure is bcc, and the average grain size is about [1], [2].
Specimen preparation, tensile tests and fractography
Experimental results
All the tensile specimens were heat treated according to the treatment described in Section 2.2. All the tensile tests reported in this article were performed on notched specimens. Whatever the chemical environment, no particular care was taken to control the oxygen activity. Three test temperatures were adopted:
(i) 20 °C for the tests performed using liquid mercury. For safety reasons we did not perform tensile tests at higher temperatures using liquid Hg.
(ii) 260 °C for the tests performed
Discussion
From the results of Fig. 2, Fig. 3, it can be seen that the notched and heat treated specimens tested in air and in liquid Hg at 20 °C undergo brittle fracture. The small differences in the mechanical response of both specimens can be attributed to differences in the notches, that we recall, are mechanically machined. Therefore, from these results, we can conclude that liquid mercury does not produce a supplementary embrittlement of the 91 steel tested in severe experimental conditions.
Conclusions
In this article we present experimental and atomic simulation results of the influence of the chemical composition of the liquid bath on the embrittlement of the grade 91 steel. Working with heat treated samples, allowed us to show the existence of an embrittlement in different liquid metals environments and to establish an experimental scale of embrittlement. Ab initio atomic scale simulations combined to a very simple thermodynamic model allowed us to build a theoretical embrittlement scale
Acknowledgements
We are indebted to P. Bocquet from Creusot–Loire Industries for providing the steel 91. This work has been supported by the French CNRS GdR GEDEON and by the EMA Department of EDF that we would like to thank. It has partially benefited from the computational facilities of the French organizations IDRIS and CINES as well as those of the CRI-USTL supported by the Fonds Européens de Développement Régional. We are also indebted to our colleague R. Besson for very fruitful scientific discussions.
References (13)
- et al.
Scripta Mater.
(2000) - et al.
J. Nucl. Mater.
(2001) - et al.
Phys. Rev. B
(1991) Coll. Surf. A
(1999)- et al.
Phys. Rev. B
(1993)et al.Phys. Rev. B
(1994)G. Kresse, J. Furthmüller, Phys. Rev. B 54 (1996)... et al.Comput. Mater. Sci.
(1996) - et al.
Phys. Rev.
(1964)et al.Phys. Rev.
(1965)
Cited by (64)
Effects of Pb and Bi on mechanical properties and fracture modes of bcc iron symmetrical tilt grain boundaries
2024, Materials Today CommunicationsA novel monitoring system for fatigue crack length of compact tensile specimen in liquid lead-bismuth eutectic
2024, Nuclear Engineering and TechnologyEffects of lead, bismuth and alloying elements on elastic anisotropy and fracture toughness of α-Fe
2023, Journal of Nuclear MaterialsFirst-principles study on the corrosion-resistant of lead-bismuth to rough stainless steel surface
2023, Journal of Nuclear Materials