Key Challenge 4
Failure Mechanisms
Leads: Fionn Dunne (Modelling) and Philipp Frankel (Experimental)
Zirconium alloys are subject to a range of extreme environmental, mechanical and thermal loading conditions in nuclear service, but components fabricated in this material must satisfy stringent design criteria, since they are often safety-critical; that is, their in-service failure potentially gives rise to major infrastructural damage and loss of life. These consequences are unacceptable, and hence there is a crucial need for the establishment of accurate mechanistically based predictive modelling techniques that enable engineers to achieve non-conservative but safe design. MIDAS’s access to neutron-irradiated Zr facilitates the detailed and rigorous extraction of properties, behaviour and performance of relevance to in-service conditions, and offers the potential to establish a transformational understanding of the degradation and failure of these materials resulting from fatigue, delayed hydride cracking (DHC), pellet-cladding interaction (PCI), and loss of coolant accident (LOCA), all of which must be avoided for highly efficient, high-burn-up next generation reactors. A major challenge exists in linking and scaling from the mechanistic understanding obtained at the material level through to the performance of reactor components. This is the subject of KC-4, and the methodology here is to utilise modelling techniques across the differing length scales, fully integrated with experimental work where possible, in order to bridge from the atomic defect and dislocation levels using accurate quantum mechanical descriptions through to molecular dynamics (MD) and DDP, to the microstructural, grain morphology, crystallography, and phase level with dislocation-based crystal plasticity (CP), and finally to component level with notches, stress raisers, thermal and mechanical loading using crystal-to-continuum homogenisation techniques. The four key scientific objectives are:
Utilise models developed in KC-2 to establish crack nucleation and growth models in Zr alloys at the microstructurally-sensitive scale under conditions of thermal and mechanical cyclic loading, with full recognition of environmental effects;
Integrate small-scale experiment (KC-2) and modelling to establish mechanistic understanding of DHC. Develop crystal-to-continuum modelling methods to give predictive capability at the component (e.g. notch) level; (Noting that these first two objectives are equally relevant to fission fuel cladding and fusion breeder blanket applications.)
Combine predictive modelling with experiments on irradiated material and detailed characterisation to provide new understanding of the mechanisms whereby fission products and mechanical stress can result in PCI during service;
Understand and establish effects of very-fast heating rates, experienced by fuel cladding during LOCA and RIA conditions, on strain patterns and failure mechanisms as a result of ballooning.
Work Packages
WP 4.1: Fatigue and Crack Propagation
The modelling method most suitable for capturing the key drivers of short crack growth and hydride interactions, and with the ability to provide efficient homogenisation to component levels, is computational crystal plasticity. This method will be demonstrated in carefully selected experiments to reflect local measured microstructural quantities obtained in KC-2. The explicit incorporation of crack growth (via XFEM – extended finite element methods) within a crystal plasticity finite element formulation has been successfully demonstrated by Imperial, providing the opportunity to model explicitly the short crack growth observed, and quantitatively characterised in experiments. Validated crystal slip rules for the Zr alloys will be established in KC-2, and here we will construct faithfully representative geometric microstructural models. In order to first simplify the problem we will start with blocky polycrystal fatigue samples that enable through thickness grains. Using a novel in-situ set-up in a FEG-SEM test station developed at the Henry Royce Institute, in conjunction with TESCAN, we will be able to monitor 2D surface slip accumulation by HR-DIC while also monitoring crack path and growth rates, with fully quantified crystallographic orientations from EBSD. In addition, HR-EBSD will be employed at Oxford for local stress and GND density analysis. We will investigate crack growth paths and growth rates experimentally and using coupled crystal plasticity and XFEM in the microstructures, evaluating mechanistic drivers against the experimental observations. Quantitative assessment will be carried out of the model and experimentally measured local slip and strain fields, GND densities, stresses ahead of the crack tips during crack growth. Through critical assessment of the integrated experiment, characterisation and model predictions, mechanistic drivers for crack path and growth rates will be inferred and the drivers validated against experimental data for initial hydride content and p-irradiated samples, before taking the experimental methodology and employing it on our n-irradiated material. The selection of samples and neutron dpa level will depend on the findings in KC-1/2 identifying specific effects of alloy chemistry and early/mid/end of life conditions.
WP 4.1 Key Deliverable
Validated, mechanistically informed microstructurally-sensitive crack growth modelling capability for Zr alloys that is sensitive to irradiation and hydrogen effects.
WP 4.2: Delayed Hydride Cracking (DHC)
Using expertise at Imperial, we will generate Zr test samples with controlled hydrogen and hydride content and utilise bespoke thermo-mechanical beam testing together with HR-DIC and EBSD, in order to investigate the mechanistic basis of DHC in laboratory observation. The HR techniques will enable us to quantify initial residual stresses generated at hydrides, and the slip development and localisation at and through the hydrides. The thermal ratcheting phenomenon will be investigated using Imperial’s high-res Questar in-situ microscope to fully quantify strain histories around loading cycles. Proton-irradiated samples will also be tested and, later in the programme, the Materials Research Facility at Culham will facilitate neutron-irradiated sample testing. The parallel discrete dislocation and crystal plasticity modelling of replicated laboratory samples will allow us contemporaneously to extract key local quantities including parameters controlling slip localisation and stress in the vicinity of hydrides, lattice curvature, GND density and stored energy. In this way we will establish the mechanistic understanding of DHC with which to inform higher-level crystal to continuum predictive modelling. The latter has recently been successfully established by Imperial for micro- to macro-hydride formation. The ability to incorporate the DHC mechanisms is potentially transformational and offers the likelihood of fast, industry-ready predictive fuel performance design codes.
WP 4.2 Key Deliverables
Mechanistic understanding of DHC, thermal ratcheting, and role of irradiation on DHC.
Mechanistic models for crack nucleation/growth and interaction with hydrides.
WP 4.3: Pellet-Cladding Interaction (PCI)
The PACE (Pellet Associated Cladding Degredation) international working group (led by MIDAS team member Philipp Frankel) has sought to combine industrial expertise with the newest simulation and characterisation techniques, in order to move away from the empirical approach that has governed PCI control. In MIDAS we will extend this work to look at the irradiated material. Methodology has been developed to relate chemical segregation around experimentally induced I-SCC cracks with local microstructure using complementary techniques. X-ray computed tomography non-destructively reveals complex crack morphology, enabling targeted high-resolution microstructural analysis. 3D characterisation of microstructure in relation to crack progression by serial sectioning using a Xe-plasma focused ion beam, is combined with high-resolution crystallographic (STEM) and chemical (NanoSIMS) mapping and will be extended to include 3DAP analysis. Complementary atomistic simulations, using accurate tight-binding models developed in KC-3, but with the inclusion of iodine/caesium, will investigate migration of these products through any inner oxide, their propensity to segregate to grain boundaries or crystallographic orientations, and the effect of these “impurities” on cleavage energies. Use of existing Zr-embedded atom empirical potentials, combined with the new fast tight- binding models for Zr metal and iodine, will enable atomistic studies of loaded crack tips and so new insights into I-SCC mechanisms. The approach maintains the chemical accuracy needed at the crack tip whilst allowing the use of >100,000 atoms to describe a crack. Comparing progression of I- SCC in different microstructures with observations from ramp-tested reactor material will produce and validate models of the influence fission products have on crack initiation and propagation.
As the presence of hydrides may have an important influence on PCI, this activity will work closely with WP 4.2 to look at combined effects of fission products, hydrides and irradiation-enhanced stress localisation. Additional in-reactor samples relevant to this WP will be available through access to the LAMDA lab based at Oak Ridge National Laboratory. For reactor samples, the recent Henry Royce Institute-funded upgrade of Manchester’s NanoSIMS with a high spatial resolution oxygen source, enables investigation of the role of low concentrations of caesium at previously unachievable high spatial resolution near crack tips.
WP 4.3 Key Deliverables
Understanding the migration and segregation of fission products in Zr alloys; relate this to their influence on crack formation and propagation.
Develop understanding of Interaction between irradiation, fission products and hydrides.
Provide a material model that accounts for PCI.
WP 4.4: Loss-of-Coolant Accident (LOCA)
During LOCA, it is critical that the integrity of the fuel cladding is maintained and release of fission gas is avoided. Predicting the mechanical performance of fuel cladding during accident scenarios is therefore critical, and can put constraints on fuel assemblies during normal operating conditions. The desire to use Zr alloys at temperatures up to 650C in fusion blankets is an additional driver to explore high-temperature mechanical performance. A key aspect to improve the accuracy of such integrity predictions for fuel cladding is to understand the phase-specific plasticity during ballooning. It is important to note that the hydrogen picked up during normal operation will segregate to the high-temperature phase when the material is heated into the two-phase regime. Such a complex process greatly affects the strength and deformation mechanisms of each phase, and is further complicated here by the biaxial stress state in a pressurised tube. To improve predictions of mechanical performance it is key to fully understand the contribution from each phase, which will be investigated by the application of imaging techniques and DIC during high-temperature deformation to record strain patterns, while in-situ loading experiments using high-energy synchrotron x-ray and neutron diffraction will provide insight into deformation mechanisms by capturing intergranular strains and using those to validate and constrain CP models. Biaxial stress conditions will be achieved by using modified sample geometries when, for example, utilising the electro-thermo-mechanical-tester based at the Diamond Light Source. The work will be carried out on hydrogen-charged material to evaluate conditions that simulate end-of-life service conditions, i.e. the worst-case scenario. In order to record strain patterns during such creep and bursting experiments 3D DIC analysis would be developed, for which the hardware already exists at Manchester.
WP 4.4 Key Deliverables
Mechanistic understanding of deformation paths at high temperatures during biaxial stress conditions.
Mechanistic understanding of the role of hydrogen on strain behaviour, particularly in the beta-phase.