Failure mechanisms in unirradiated ZIRLO® cladding with radial hydrides

https://doi.org/10.1016/j.jnucmat.2020.152668Get rights and content

HIGHLIGHTS

  • Radial hydride metrics is not the only parameter for cladding failure

  • Crack nucleation occurs in radial hydrides at regions with highest hoop stress

  • The failure of empty pre-hydrided cladding with radial hydrides is weakest-link type

  • Unstable crack propagation occurs at 0.5 mm displacement in the ring compression test.

Abstract

The purpose of this paper is to investigate the relationship between hydride morphology, in particular the presence of radial hydrides (RHs), stress state and failure mechanisms associated with the ring compression test (RCT). Samples of ZIRLO® cladding were pre-hydrided and subjected to thermo-mechanical treatments to precipitate long radial hydrides. The results show that the reorientation treatment was very successful. A considerable fraction of RHs was generated, the radial hydride continuity factor being around 80 to 90% of the wall thickness. The samples with reoriented hydrides were tested using the RCT at room temperature. Macroscopic brittle failure was observed with sudden load drops for displacements around 0.5 mm, with a calculated “offset strain” between 0.5 and 1%. Crack nucleation occurs in RHs located in regions with the highest values of hoop stress. These locations are the inner diameter of cladding at the vertical plane of the sample (12 and 6 o'clock positions) and the outer diameter at the horizontal plane (3 and 9 o'clock positions). Noticeable load drops in the RCT are associated with unstable crack propagation events through the radial hydride network, the crack front reaching up to 90% of the wall thickness in some cases. The failure micro-mechanism is quasi-cleavage in the hydrides and micro-void nucleation, growth and coalescence in the Zr matrix, with ductile tearing patches connecting neighboring hydrides. The main conclusion is that radial hydride metrics is not the only parameter that determines cladding failure in the presence of RHs, but the interaction between the location and continuity of RHs and the stress normal to the hydride (the hoop stress in this case). Consequently, if a radial hydride is located at a position within the cladding where the hoop stress is small, a crack will not be initiated easily in the RCT.

Introduction

Zirconium hydrides are key to understand the failure mechanisms of spent nuclear fuel zirconium-alloy cladding in light water reactors (LWR) because of their brittleness when compared to the Zr matrix.

Zirconium alloys are used as nuclear fuel cladding tubes in LWR due to a combination of corrosion resistance, neutron transparency and mechanical properties at operating temperatures. In the primary circuit water, the corrosion reaction will produce zirconium oxide and hydrogen at the outer surface of the tube, both processes threatening the cladding integrity [1].

The hydrogen pickup fraction –the relative amount that will be picked up inside the cladding alloy– under normal pressure water reactor (PWR) operation is comparable for Zircaloy-4 and ZIRLO® (approximately 15%) [2]. It seems that hydrogen enters the oxide layer in the form of protons and diffuses inside the zirconium matrix, where it can be in solid solution or, if the hydrogen concentration is high enough, will precipitate as zirconium hydride [1]. Hydrides will mainly precipitate along the circumferential direction of cladding during reactor operation in cold-work stress-relieved (CWSR) cladding, whereas in recrystallized alloys a mixture of circumferential and radial hydrides is usually found [3]. The hydride morphology resulting from operation does not change when spent fuel is maintained in the storage pool.

This may not be the case when dry storage is implemented. During vacuum drying prior to the storage period, the cladding temperature will increase due to the absence of coolant; the maximum temperature recommended by the NRC being 400°C [4]. As the terminal solid solubility for dissolution (TSSD) of hydrogen in zirconium at this temperature is approximately 200 ppm (see [3] for a review of data obtained from differential scanning calorimetry and synchrotron diffraction), a portion of the hydrides will re-dissolve in the matrix. The internal pressure in the rod will increase with temperature from the end of life value (around 4 MPa for standard PWR rods at 25°C [5]); the external pressure being negligible in this case –with respect to the one during operation inside the reactor. After vacuum drying, the cladding temperature will slowly decrease from its maximum value during the dry storage phase, and so will the internal pressure. Consequently, a fraction of the hydrogen previously dissolved will re-precipitate when the terminal solid solubility for precipitation (TSSP) is reached. For a beginning of storage temperature of 400°C, it has been reported that about 6.5 years of dry storage are needed for the temperature to drop to the hydride precipitation solvus of 335°C [5]. If the hoop stress due to internal pressure is large enough, a portion of the newly precipitated hydrides will find it easier to grow perpendicular to the hoop stress, i.e. along the radial direction. This phenomenon is known as hydride reorientation. It was first documented in the 1960’s for Zircaloy-2 [6], [7] and has been extensively studied, particularly for Zircaloy-4 [8], [9], [10], [11], [12], [13], but also for ZIRLO® and other zirconium alloys [10], [14], [15].

It has been demonstrated that hydride reorientation may degrade the cladding ductility due to the brittleness of the zirconium hydrides [6], [10], [14], [15]. As cladding is the first structural barrier to contain the fission products (both gases and solids), it is crucial to understand this phenomenon, to guarantee its structural integrity in dry storage conditions. A considerable effort has been made to purposely precipitate hydrides along the radial direction of cladding, both in irradiated and unirradiated cladding. Most works have focused on the importance of hydride morphology, and several metrics have been proposed to define the degree of hydrogen reorientation, like the radial hydride fraction (RHF) [16] and the radial hydride continuity factor (RHCF) [10]. Another definition of RHF is proposed in [17], although it is recognized that it cannot capture some key features related to the connectivity between closely located hydrides [3] that are better represented by the RHCF.

Less attention has been devoted to the precise relationship between the hydride morphology and the stress state in the cladding, the latter being dependent on the postulated operation or accident conditions. The failure of fuel rods under hypothetical transportation accidents was thoroughly studied in [18], and three possible failure modes were identified. Only one of them, designated as Mode-III, is affected by RH, and is totally decoupled from the other two modes [8]. In this failure mode, a cladding fracture may occur in the longitudinal direction of the tube under a compressive (pinch) load as a consequence of a postulated 9-meter cask drop accident [18].

In laboratory, this failure mode can be represented by the ring compression test (RCT), which is used to characterize the mechanical response of a small cladding sample taken from the fuel rod [10], [19]. This test is easy to perform but is far from a standard tensile test regarding interpretation. Consequently, it is important to understand precisely the stress state in the cladding wall and its evolution with applied compressive load in order to avoid misconceptions. The stress state associated with the RCT is not homogeneous in the cladding cross-section, with bending deformations of opposite sign in the generatrices corresponding to the equatorial (3 and 9 o'clock positions) and vertical (12 and 6 o'clock positions) diameters of the sample. The maximum stresses occur at the inner diameter (12 and 6 o'clock positions) and at the outer diameter (3 and 9 o'clock positions) of cladding, respectively. The hoop stress will be comparatively low for a radial hydride located at the mid-wall section (even in the generatrices where maximum stresses occur) and consequently, a crack will not be initiated easily at this location. Consequently, not only the hydride orientation, but also its length and particularly its radial position will determine the structural integrity of cladding in the RCT. This problem is different to the one that has usually been considered to study the effect of hydrides in the reduction of ductility of zirconium alloys, because the stress state is more complicated. Several models have been developed by considering a remote tensile stress and assuming that cracks will initiate in the hydrides and will propagate by plastic deformation through the matrix ligaments located between the hydride micro-cracks. It has been reported that this process will depend on the size, volume fraction, morphology and distribution of hydrides, as well as the continuity of the hydride network and the relationship between the orientation of the hydride platelet and the stress [20], [21].

The main objective of this paper is to correlate the failure mechanisms with hydride morphology and stress state in pre-hydrided samples with RHs. To this end, cylindrical ring samples of unirradiated ZIRLO® cladding were pre-hydrided and subjected to a carefully controlled hydride reorientation treatment (RHT) at a constant hoop stress of 140 MPa. This is not representative of drying-storage conditions but is aimed at obtaining a significant fraction of long reoriented hydrides in the radial direction. Finite element analysis (FEA) was performed to determine precisely the loads to be applied in the RHT and to characterize the stress state associated with the ring compression test. Pre-hydrided samples after the RHT were subjected to RCT at 20°C. From the experimental results, along with optical and SEM microscopy observations, the relation between the microstructure, the failure micro-mechanisms, and the stress field during RCT was established.

Section snippets

Material

Samples were cut from ZIRLO® cladding tubes, with 9.5 mm outer diameter and 0.57 mm wall thickness. The material is cold work stress-relieved and was provided by ENUSA Industrias Avanzadas, S.A.

ZIRLO® is a zirconium alloy widely used as cladding material in light water reactors (LWR). It was developed as an improvement of Zircaloy-4 against corrosion. The alloy composition is given in Table 1.

Hydrogen charging

Hydrogen was introduced in 15-mm high cylindrical cladding samples by cathodic charging. The inner

Cathodic charging of samples

Table 4 shows the hydrogen content of three samples (RCT-01, RCT-02 and RCT-04), measured by the hot extraction technique at two different locations in each sample.

Radial hydride treatment

Fig. 5 corresponds to the cross-section of sample PT-03 after hydrogen charging and precipitation treatment. Figs. 6 and 7 show the cross-sections at 12 and 3 o'clock positions, respectively, of sample RHT-03 after the hydride reorientation treatment. The thin, black lines are the zirconium hydrides revealed by chemical etching. For

Hydrogen charging and hydride reorientation treatment

At the maximum temperature employed in the hydride reorientation treatment (400°C in this case), the terminal solid solubility for dissolution (TSSD) for ZIRLO® is around 230 ppm (calculated from the results recently reported in [27]). As shown in Table 4, the hydrogen content was fairly close to the target value (150 ppm), with an average value of 163 ppm of hydrogen and a small standard deviation (+/- 18 ppm), which proves the reproducibility of the experimental technique employed for

Conclusions

Hydrogen was introduced in samples of unirradiated ZIRLO® cladding tubes (9.5 mm outer diameter and 0.57 mm wall thickness) by means of cathodic charging, the average measured content being 163 ± 18 wppm.

After hydrogen charging, samples were thermally treated to homogeneously distribute the hydrogen and to precipitate zirconium hydrides along the hoop direction. Then the samples were subjected to a thermomechanical treatment with the aim of reorienting the hydrides in the radial direction (RHT)

CRediT authorship contribution statement

J. Ruiz-Hervias: Conceptualization, Methodology, Investigation, Writing - review & editing, Supervision, Project administration, Funding acquisition. K. Simbruner: Methodology, Software, Investigation, Writing - review & editing, Visualization. M. Cristobal-Beneyto: Methodology, Investigation, Writing - review & editing, Visualization. D. Perez-Gallego: Methodology, Investigation, Visualization. U. Zencker: Conceptualization, Software, Methodology, Writing - review & editing, Supervision,

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Funding: This work was supported by BRUZL research project by the German Federal Ministry for Economic Affairs and Energy (BMWi) [contract No 1501561]; the European Union's Horizon 2020 research and innovation programme [agreement No 847593 (EURAD project)]; and the Spanish Ministry of Science and Innovation and Universities through [grants number RTI2018-097221-B-I00 and PGC2018-097116-A-I00]. ENUSA Industrias Avanzadas supplied the material for this research, and the authors are indebted to

References (28)

  • F.J. Gómez Sánchez et al.

    A new procedure to calculate the constitutive equation of nuclear fuel cladding from ring compression tests

    Prog. Nucl. Energy

    (2017)
  • M.A. Martín-Rengel et al.

    Revisiting the method to obtain the mechanical properties of hydrided fuel cladding in the hoop direction

    J. Nucl. Mater.

    (2012)
  • A.T. Motta et al.

    Corrosion of Zirconium Alloys Used for Nuclear Fuel Cladding

    Annu. Rev. Mater. Res.

    (2015)
  • A.M. Garde et al.

    Hydrogen Pick-Up Fraction for ZIRLO™ Cladding Corrosion and Resulting Impact on the Cladding Integrity

  • Cited by (3)

    View full text