Abstract
Zirconium alloys are widely used in the nuclear industry as structural materials and as cladding for nuclear fuel in fuel assemblies. However, zirconium and its alloys have a considerable affinity for hydrogen, which readily diffuses into zirconium at high temperatures but has low solubility in the hexagonal α-phase. During operation in a boiling or pressurized water reactor, the cladding material will undergo aqueous corrosion. Some of the hydrogen which is produced during such corrosion is picked up by the zirconium cladding material and will lead to hydride precipitation once the solubility limit has been exceeded. As there is a temperature gradient between the inner surface (fuel = hot) and the outer surface (coolant = less hot) of the cladding, a hydride rim is typical after some operation time. It is well known that the presence and orientation of these hydrides adversely affect the mechanical properties of the material [1], [2] and [3]. They can lead to embrittlement, delayed hydride cracking (DHC) and hydride blistering, all of which reduce the lifetime of the component and are cause for considerable environmental concern in the storage of spent fuel rods [1]. The widely accepted mechanism for DHC assumes that hydrogen diffuses along the stress gradients towards the tensile crack tip area where (re)precipitation of the brittle hydrides occurs, which at certain hydride size and stress intensity then encourages further propagation of the crack. Hydrogen can occupy interstitial tetrahedral and octahedral sites in hcp crystals, but for zirconium hydrides, the reported crystal structures from diffraction measurements indicate that mainly the tetrahedral sites are occupied. Two stable hydride phases, δ and ε, and one “metastable” hydride phase, γ, have been reported in the literature. The exact nature of the metastable phase remains controversial [4] and [5], and it is noteworthy that the reported stable δ-phase and metastable γ-phase essentially only differ in the degree of ordering of the hydrogen atoms on tetrahedral sites, which in turn affects the composition (see section 5). Yet, to the best of the authors’ knowledge, no direct observation of a transition between the “ordered” and “unordered” states of the room temperature phases has been reported in the literature. Furthermore, the elusive metastable γ-phase is observed mostly after rapid quenching of the sample (e.g., >10 K min–1) [3]. Improving understanding of the behavior and properties of these zirconium hydrides is therefore of significant importance. The hydride bands can be easily revealed by conventional laboratory-based imaging techniques, and it is well known that hydrides change orientation, depending on stresses [3]. However, these techniques are essentially surface techniques and provide very little information about the micromechanical properties of the hydrides, their crystallography and their relationship with the matrix [6] in the bulk, where the additional constraints are likely to change any transformation properties with respect to those occurring at the surface. This problem has been recognized, for example, in the study of martensitic transformation, to which the hydride transformation in zirconium alloys bears some resemblance [7] and [8].
Non-destructive characterization techniques such as neutron and high-energy synchrotron X-ray diffraction provide the capability to investigate phase-specific mechanical properties in bulk materials. For neutron diffraction, in order to overcome the large incoherent neutron scattering cross section of hydrogen, it is often replaced by its isotope deuterium [9]. Synchrotron X-rays, in contrast, are becoming increasingly popular for structural investigations of engineering materials, as third-generation synchrotron X-ray sources, such as the ESRF in Grenoble, France, yield an extremely high X-ray flux at energies high enough to penetrate metals. Unlike laboratory X-ray sources, this allows non-destructive investigations in bulk metallic components with a penetration depth of several millimeters, or even centimeters in light alloys at very high spatial resolution. Similarly, the large flux opens up the possibility of performing in situ mechanical testing with adequate time resolution, even in materials with a relatively high atomic number such as zirconium and on phases with a low volume fraction such as hydrides in zirconium.
The aim of this experiment was to undertake a uniaxial tensile test on hydrided zircaloy-2 and zircaloy-4 samples in order to study in situ the elastic response of different diffraction peaks (lattice planes) of both the matrix and the hydride phase under load and during plastic deformation. Such observations are expected to improve the understanding of the deformation and fracture mechanisms of hydrided zirconium alloys and provide valuable verification data for modeling approaches of failure mechanisms.
The experiment was performed using the high-energy beam line ID15A (ESRF) in energy-dispersive mode. The results revealed the anticipated response of the zirconium matrix, but showed surprisingly large d-spacing shifts of the hydride peaks incompatible with elasticity theory. The observed hydride peak shift can be explained by a stress-induced transformation of the hydride phase due to gradual ordering of the hydrogen atoms on the tetrahedral sites in the subset of grains most closely aligned with the loading direction, which is discussed here. Subsequent annealing resulted in a reverse hydride peak shift, which suggests reversal of the ordering process.
Translated title of the contribution | Evidence of stress-induced hydrogen ordering in zirconium hydrides |
---|---|
Original language | English |
Pages (from-to) | 145 - 152 |
Number of pages | 8 |
Journal | Acta Materialia |
Volume | 57:1 |
DOIs | |
Publication status | Published - Jan 2009 |