Oxidation of AISI 348 Stainless Steel in Water at High Temperatures

Introduction: The Fukushima Daiichi Nuclear Plant accident in March 2011 brought great challenge on the development of Accident Tolerant Fuels (ATF). Even though austenitic steel fuel cladding cannot be considered as an ATF, this material operated reliably in the first Pressurized Water Reactors (PWR). Despite being favoured over stainless steel from the standpoint of neutron economy, zirconium-water reaction generates substantially more energy than stainless steel. Also, due to its higher oxidation resistance up to 1200 °C, stainless steel cladding may present some safety advantages.

Methods: Oxidation of AISI 348 stainless steel was investigated in water at 1000 – 1350 °C by Thermal Gravimetric Analysis (TGA). Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDS) were applied for samples characterization.

Results: Linear-Parabolic kinetics and multilayer oxide scales and voids were found. Based on the experimental results, AISI 348 type presented a higher activation energy comparing to Zircaloy-4, which was also tested for validation and performance comparison.

Discussion: Stainless steel cladding provides a safety improvement when applied as cladding by means of the reduction of the heat sources due to its smaller exothermic metal-water reaction compared to zirconium-based alloys; consequently, the use of AISI 348 as cladding material instead of Zircaloy-4 would provide a slower accident progression, for a Small Modular Reactor (SMR).

The U-Gd-O system has several technological applications involving a wide range of composition which makes it important to characterize its crystallographic structures in all composition ranges [1,2]. However, there is a specific range, Gd / U> 1, in which there is no agreement between the published results [3]. In this range, Vegard's Law is not obeyed and the possible existence of multifaces has not been well understood [4]. The samples were characterized by SEM and X-ray diffraction. Rietveld refinement and EDS maps were applied to a partially disordered crystallographic state, providing the interpretation of the crystalline phases. The method allows the characterization of a biphasic region where the X-ray diffractogram apparently shows to be a single phase. The batch of samples were prepared following coprecipitation from DUA route with Gd2O3 content in the 50 to 90 wt%. The batch of samples were sintered in a dilatometer up to 1650 °C for 3 hours. The X-rays diffractograms of U-Gd-O powders were obtained from the half of sample pellets pulverized in agar gral using a conventional X-ray sources. Rietveld analysis was applied to obtain cell parameters, atomic positions and atomic displacement factors and compared with literature available of each phase identified [5]. Also, the quantification of phases was performed for the different contents of Gd2O3 in the system. The other part was used as built-in SEM section. SEM captures were done using a FEI Microscope, type QUANTA 600 FEG. Energy Dispersive Scattering (EDS) measurements were conducted by an EDS Oxford Instruments.

1. M. Durazzo, H.G. Riella, IAEA Advanced Fuel Pellet Materials and Fuel Rod Design for Water Cooled Reactors (2009) 35.

2. Jena, Hrudananda, et al. "X-ray powder diffraction of RE6UO12 (RE=Eu, Gd, and Dy)" Powder Diffraction 16.4 (2001): 220-223.

3. D. Pieck, L. Desgranges, P. Matheron, H. Palancher, “Evidence of a new crystalline phase in U-Gd-O phase diagram” Journal of Nuclear Materials, Volume 461, pp.186-192 (2015).

4. Durazzo, M., Oliveira, F. B. V., de Carvalho, E. U., & Riella, H. G. (2010). “Phase studies in the UO2–Gd2O3 system”. Journal of Nuclear Materials, 400(3), 183-188.

5. Young, R.A. (Ed.), “The Rietveld Method, International Union of Crystallography” Oxford University Press, Great Britain, 1993.

(U, Pu)O₂ MOX fuels represent a considerable fraction of the nuclear fuel used in operating nuclear power reactors [1]. MOX are also considered as a potential fuel for Gen-IV sodium fast reactors (SFR) [2], with a relatively higher Pu content. The thermophysical properties of actinide oxides have been extensively investigated in the past decades [3, 4]. An accurate understanding of MOX properties is crucial for fuel performance. For instance, disentangling the effect of the mixture is the key to accurately estimate the role of minor actinides in the fuel properties. However, the origin of some behaviours such as the Neumann-Kopp rule for the specific heat and the decrease in the thermal conductivity of actinide mixed oxides are still not fully understood. In the present study, we have applied Hubbard corrected density functional theory (DFT+U) and empirical potentials coupled to the Boltzmann transport equation (BTE) to investigate the harmonic and anharmonic lattice vibrational properties of MOX fuels [5]. The specific heat, entropy and thermal conductivity, were investigated. We demonstrate that the gaps in the specific heat of actinide oxides, and thus the Neumann-Kopp rule, originate from the contribution of the crystal electric field at the actinide cations.

References

[1] Provost et al, MOX and UOX PWR fuel performances EDF operating experience, J. Nucl. Sci. Tech 43 (2006) 960–962.

[2] Vaudez et al, A new fabrication route for SFR fuel using (U, Pu)O₂ powder obtained by oxalic co-conversion, J. Nucl. Mater 442 (2013) 227–234.

[3] Carbajo et al, A review of the thermophysical properties of MOX and UO₂ fuels, J. Nucl. Mater 299 (2001) 181–198.

[4] Fink, Enthalpy and heat capacity of the actinide oxides, Int. J. Thermophys 3 (1982) 165–200.

[5] Njifon et al, Phonons and thermophysical properties of U_1−yPu_yO₂ mixed oxide (MOX) fuels, J. Nucl. Mater, Under review.

TRISO fuel modelling results using the finite-element nuclear fuel performance code Bison coupled with the equilibrium thermodynamics solver code Thermochimica will be presented. A thermodynamics database detailing the behaviour of fuel and fission product elements including Pu, U, Nd, Pr, Ce, La, Ba, Cs, Xe, I, Te, Ag, Pd, Rh, Ru, Tc, Mo, Zr, Y, Rb, Ar, O, and C is employed. Of particular interest will be the fission product behaviour in TRISO fuel kernels and transport to surrounding layers. Pd is of special concern as it attacks the SiC containment layer of the TRISO particle.

We investigated the heat capacity and thermal conductivity of UN using first principles methods. The generalized gradient approximation of the Perdew, Burke, and Ernzerhof (/PBEsol) as implemented in Quantum ESPRESSO (QE) and associated codes (EPW, Boltztrap, ShengBTE), was used.  We evaluated the energy of the UN to be lower for ferromagnetic ordering than non-magnetic. We found, using QE code, that the lattice constant calculated using PBEsol functional and norm-conserving pseudopotentials is slightly larger for ferromagnetic UN (0.497 nm) than non-magnetic UN (0.489 nm) and they agree with experiment. We noted a significant contribution from the optical phonons to the lattice thermal conductivity, which was previously observed experimentally for urania. The phonons’ calculated contribution to the thermal conductivity, which decreases with temperature, is smaller at room temperature (7.20 Wm^-1K^-1) than evaluated from the correlation recommended for urania (8.79 Wm^-1K^-1). However, urania’s thermal conductivity deteriorates faster with temperature; therefore it becomes lower than that calculated for UN for temperatures higher than 490 K. The total thermal conductivity, evaluated here, leads to the total thermal conductivity of UN being overestimated below 1000 K. Therefore, further investigations are needed to evaluate the effect of the interaction of magnetic moments on uranium with phonons and electrons. Furthermore, the electronic thermal conductivity can only be performed for non-magnetic UN, using EPW and Boltztrap codes. The results are dependent on a selection of electronic carriers but show good qualitative agreement with experiment at higher temperatures. The calculated electrical resistivity of the UN at low temperature is much lower than measured, but is similar to the experimentally measured behaviour of non-magnetic ThN. On the other hand, Ziman’s model predicts two orders of magnitude lower resistivity than presently calculated, which is in strong disagreement with the measured resistivity of UN, but compares well with the resistivity of aluminum.

Defect generation, micro-chemical changes and changes to mechanical properties as a function of fluence has been well studied and characterized for austenitic stainless steels subjected to LWR operating temperature (~280-350^oC). Additionally, microstructural and mechanical changes from irradiation damage, i.e. RIS and increased hardness, have been identified as potential important parameters affecting the susceptibility of irradiation assisted stress corrosion cracking (IASCC). Although, the CANDU reactor design predominantly employs Zr-alloy components in-core, a key austenitic stainless steel component, the calandria vessel containing the moderator, is located at the periphery to the core. The vessel is subjected to a high thermal neutron flux unique to the CANDU reactor, albeit at a significantly lower flux than LWRs. Recently, in the context of long-term operation for periods as long as 100 years, gaps in knowledge were identified pertaining to the effects irradiation on the mechanical properties and IASCC susceptibility of the CANDU calandria vessel and its welds. In parallel with industry efforts to obtain neutron-irradiated test materials, the suitability of proton irradiation is investigated as a means to interrogate irradiation damage under the somewhat unique environmental conditions of the calandria vessel, such as its low operating temperature. A comparative irradiation damage study, employing both low (100^oC) temperature proton irradiation was carried out on 304L, 316 and 308 stainless steel materials. TEM/EDS and APT analyses were performed to characterize irradiation-induced changes in microstructure, and nanoindentation was used to probe changes in mechanical properties.  

Coated particle fuels built around uranium oxycarbide or uranium nitride kernels are of interest for several reactor applications given their ability to retain fission products under a range of conditions. The most common modern coated particle fuels (e.g. TRISO) consist of a SiC layer between inner and outer pyrolytic carbon layers. Prior studies of TRISO performance have found that several mechanisms can affect the integrity of the SiC layer. In particular, SiC loses its mechanical strength above 1700°C, can be rapidly consumed in a very short amount of time forming SiO(g) if the IPyC layer fails, and is vulnerable to localized attack by fission products such as palladium. In order to raise the possible operating temperature and avoid these issues, alternative coatings such as ZrC and TiN are being examined. While the performance of ZrC under irradiation has received previous research, there is generally a lack of information surrounding nitride-based coatings and their behavior under irradiation. Two key aspects of particular interest for the research community are the evolution of the thermal conductivity and the volumetric swelling after irradiation. Ion irradiation effects on Si₃N₄ and ZrN properties and microstructure were investigated in this work.

ZrN samples were sintered using SPS while Si₃N₄ was obtained from a commercially available source. Samples were irradiated using 15 MeV Ni⁵⁺ ions at midrange doses varying from 1 to 50 dpa and temperatures from 300 °C to 700 °C. These temperatures and doses are within the range anticipated for operation of TRISO fuels in light water reactors and will also bring us important behavioral information at lower temperatures, preceding future work at higher temperatures targeted towards small modular reactors, reactor for space nuclear power, and other applications. Ion irradiation was selected to simulate radiation-induced displacement damage defects from in-pile neutron irradiation without the cost, time and technical challenges attached to them. Post irradiation characterization consisted of grazing incidence X-ray diffraction, nanoindentation as well as scanning and transmission electron microscopy. Ion irradiation effects observed on both ZrN and Si₃N₄ will be presented. 

Chlorine-induced stress corrosion cracking of stainless steel 304 was investigated in a simulated marine environment using synchrotron x-ray tomography to track the progression of cracks. Fatigue pre-cracked samples in air were obtained and smaller single edge crack samples were machined out using electrical discharge machining. A custom made testing machine was built which allowed to hold a load on to the samples as well as control the temperature and relative humidity inside a kapton chamber used to simulate marine environment. Magnesium chloride solution was added to the crack. X-ray tomography data was collected in-situ at regular intervals of time throughout the experiment and the results showed the progression of the crack into a branching geometry near the surface. Finite element analysis and stress intensity analysis were conducted to explain the observed crack branching. In addition, Scanning and Transmission Electron Microscopy analysis equipped with Energy Dispersive Spectroscopy (EDS) helped evidence the presence of CrCl2, NiCl2·6H2O, and FeCl2 as well as the expected Fe3O4, Fe2O3, and Cr2O3 and the chemical maps showed the ingress of chlorine while the magnesium stayed on the peripheral of the crack.

Introduction.   Dissimilar metal welds in LWRs are susceptible to degradation via IASCC.  We performed irradiation-assisted stress corrosion cracking (IASCC) measurements of a 304L-508LAS weldment as part of the US DOE NEUP LWR sustainability program.  The 308L weld filler material recrystallizes during the welding process, with both delta ferrite and gamma austenite forming.  This mixed phase microstructure prevents hot cracking of the weld during cooling. 

Methods.  A nuclear grade 304L-308L-309L-508LAS weldment was fabricated EPRI.  The 308L filler material and 309L butter material were irradiated at 360 C with 2 MeV protons to 5 dpa.  SEM, TEM, and STEM-EDS were used to analyse the water facing irradiated and unirradiated surfaces after hot water immersion testing in NWC conditions (290 C, 10 MPa, 2000 ppb DO).  In situ cracking was initiated by applying a tensile load to approximately 5% strain at a strain rate of 1x10-7 1/s.  Lift outs were fabricating using a FEI Helios 600I dual beam FIB.  The STEM-EDS images were obtained using a FEI Themis Z electron microscope.

Results and discussion.   Tensile engineering stress-strain curves for unirradiated specimens that spanned the entire weldment are shown in the first figure for different environmental factors, including Samples 4 and 6 in NWC (2000 ppb DO).   A 3X reduction in strain rate (Sample 4 to 6) induced failure at significantly lower total strain due to longer environmental exposure.  Proton irradiation induced significant cracking under the environmental conditions specified in the Methods section compared to unirradiated material from the same specimen.  A STEM image of a FIB lift out at a crack in irradiated 308L filler material orthogonal to the free surface is shown in the second figure.   Localized deformation channels are evident in the STEM image.   The elemental map (second figure inset) indicates changes in composition that correlate to localized slip.

Low enriched uranium-molybdenum (U-Mo) fuel is being investigated as a replacement for high enriched uranium fuels in high-performance research and test reactors to minimize proliferation of nuclear materials. The behavior of U-Mo fuels in reactor environments is informed by the microstructural evolution of the fuel, which dictates its performance and lifetime in the reactor. The formation, growth and interconnection of fission gas pores contributes to the release of fission gases from the fuel meat to the fuel cladding resulting in swelling, delamination, pillowing and potential failure. This work developed and tested a tool that investigates the degree of the interconnectivity of fission gas pores irradiated U-Mo fuels. The calculated fission gas pore interconnectivity was found to be positively correlated to the fission density as well as the fission gas pore morphology. The results showed that the rate of increase of the porosity with fission density is almost 4× the rate of fission gas pore interconnectivity at the fission densities studied (~4-6x10²¹fissions/cm³). This evidence reveals that as the fission gas pores form and grow, they do not become interconnected immediately. These findings can help inform computational models to predict fuel behavior in reactor environments.

In the reactor, nuclear fuels are subjected to harsh environments that can change both the structural and chemical nature of the fuel. Fission leads to the production of new phases and precipitates that can have different, undesirable properties compared to the bulk of the fuel. Large radial thermal gradients result in a non-uniform microstructure across the pellet where grain sizes can be vastly different in regions separated by only a few millimeters. Understanding the evolution of both the microstructural and chemical nature of the fuel is of particular importance because of their effect on fuel performance and ultimately the safety of the reactor. In this work, focused ion beam (FIB) tomography, complementary electron backscatter diffraction (EBSD), and energy dispersive X-ray spectroscopy (EDS) are used to characterize the radial microstructural and microchemical evolution in high burnup oxide fuel. Coupling FIB tomography with these techniques will facilitate three-dimensional (3D) characterization of each aspect to provide a comprehensive understanding of the microstructure.

Effect of UV Irradiation on in-situ Electrochemical Behavior of Zirconium Alloy at High Temperature Water Conditions

Taeho Kim¹, Zefeng Yu², Mohamed Elbakhshwan¹, Antoine Ambard³, and Adrien Couet^1,2*

¹Department of Engineering Physics, University of Wisconsin-Madison, Madison, WI 53711, USA

²Department of Material Science and Engineering, University of Wisconsin-Madison, Madison, WI 53711, USA

³ EDF Research and Development, Materials and Mechanics of Components, Ecuelles, 77818 Moret-sur-Loing, France

^*Corresponding author: couet@wisc.edu

Abstract

The effects of UV irradiation on in-situ electrochemical behavior of oxide films on Zircaloy-4 have been investigated in this study. The objective is to investigate the effect of Cerenkov radiation on electrochemical response of fuel cladding. The temperature and pressure are controlled at 320 ℃ and 13 MPa, respectively using a recirculatory loop, and the sample is irradiated with a 200 mW/cm² UV source through a sapphire window while in the autoclave. The electrochemical behavior of Zircaloy-4 under UV is investigated with open circuit potential, potentiostatic and dynamic polarization, and electrochemical impedance spectroscopy. Fe oxide rich deposits are observed on the top of zirconium oxide surface under UV irradiation, and the kinetics of zirconium oxide growth are altered under UV irradiation. Scanning Electron Microscopy, and Transmission Electron Microscopy are used to characterize the deposits and the oxide layer. The electrochemical data including Tafel behavior, oxide resistivity and impedance values for irradiated UV and non-irradiated UV zirconium oxide is analyzed to better characterize the effects of UV on oxide properties under UV. A photo-electrochemical mechanism is proposed to explain the Fe oxide deposition and the effect of UV on corrosion mechanism and Fe oxide deposits nucleation are also discussed.

Introduction.  Neutron radiography (nR) and neutron computed tomography (nCT) are ideal non-destructive probes of hydrogen containing metals and can render direct images of hydride structures (as opposed to reciprocal-based techniques).  The large incoherent scattering cross section of hydrogen leads to significant contrast and image quality when strong neutron sources are used.   Hydrides in Zr-based LWR cladding metals lead to embrittlement, significant stored hydrogen inventory, and the potential for delayed hydride cracking during long-term storage.

Methods.  Blocky surface hydrides were created on Zircaloy 2 and 4 tubing and bar stock by exposure to H2 gas at 400 C.  In the case of hydride lenses or blisters, the tubing was masked prior to Ni evaporation so that only a round portion of the tube surface was coated.  The Ni causes an increase in the hydrogen chemical potential.  The nCT was performed using the ORNL HFIR CG-1D Cold Neutron Imaging Facility and a neutron scintillator camera.  

Results and discussion.  We show two sets of samples in the figures below, a “city scape” of samples with blocky surface hydrides in the first figure and an individual cladding tube with a hydride lens in the second figure.  Both figures show the sample set image together with nCT image(s).  The absolute hydrogen concentration can be calibrated by correlating the known total hydrogen concentration absorbed by the sample during hydrogen charging to the neutron transmission measured in the nCT experiment.

The influence of swift heavy ion (SHI) irradiation on the microstructure of polycrystalline SiC individually implanted with silver (Ag) and strontium (Sr) were investigated using Raman spectroscopy, transmission electron microscopy (TEM), scanning electron microscopy (SEM) and Rutherford backscattering spectrometry (RBS). 360 keV silver (Ag) and strontium (Sr) ions were implanted separately into SiC at room temperature (RT) all to a fluence of 2×10¹⁶ cm^-2. Some of the implanted samples were irradiated with Xe ions of 167 MeV (SHI) to a fluence of 3.4×10¹⁴ cm^-2 and 8.4×10¹⁴ cm^-2, all at RT. The as-implanted and SHI irradiated samples were vacuum annealed from 1100 to 1500 ^oC in steps of 100 ^oC for 5 hours. Implantation of silver (Ag) and strontium (Sr) amorphized the SiC, while SHIs irradiation of the as-implanted SiC resulted in limited recrystallization of the initially amorphized SiC. Annealing at 1100 °C caused more recrystallization on the un-irradiated but implanted samples compared to SHI irradiated samples. This poor recrystallization of the irradiated SiC samples was due to the amount of impurities (i.e. concentration of Ag or Sr atoms) retained after annealing at 1100 ^oC. Raman and SEM results showed that annealing of the un-irradiated but implanted samples at 1100 °C resulted in large average crystal size compared to the irradiated samples annealed in the same conditions. At 1500 ^oC, a carbon layer appeared on the surface of the irradiated and un-irradiated but implanted with Ag and Sr samples. This was due to the decomposition of the SiC and subsequent sublimation of silicon leaving a free carbon layer on the surface.

The goal of this study is to describe accurately structural, electronic and magnetic ground-state properties of UO₂, using ab initio electronic structure calculations. Experimental assessment of the magnetic ground-state of UO₂ at low temperature is challenging because of complications posed by radioactive decay, the tendency to oxidize and form non-stochiometric UO_2+x, and the toxicity of the material. Computational methods do not suffer from these difficulties, and can provide complementary critical information on the material. What is known from experiments is that UO₂ is an antiferromagnet with a fluorite-type (Fm3̄m) structure. Below the Néel temperature, however, small oxygen displacements arise, lowering the symmetry. Previous DFT studies had found the magnetic ground state of UO₂ to be (collinear) transverse 1k anti-ferromagnetic, which is not  in line with the Pa3̄ crystal symmetry suggested by some experiments. ^1,2

This work focuses on ground-state properties of UO₂ and on benchmarking the DFT calculations. We discuss the functionals and parameters needed to describe UO₂ accurately and research the magnetic ground-state of the material. We’ve established that using PBE+U and HSE06 (using the effective U value derived from experiment, ³ and standard α = 0.25) functionals to describe the localised f-states is not enough to create the correct (Mott-)insulating behaviour of the material when it has a non-collinear anti-ferromagnetic ordering. The inclusion of spin-orbit coupling is necessary to shift the f-levels and open the band gap. We also determine the screening parameter α in HSE06 that better describes the effective screening in UO₂ and its electronic properties.

¹ Desgranges L et al 2017 Inorg. Chem. 56 321−326
² Faber J and Lander GH 1975 Phys. Rev. Lett. 35 1770
³ Kotani A and Yamazaki T 1992 Prog. Theor. Phys. Suppl. 108 117

The objective of this work is to acquire data on the fission products behaviour during the different stages of a Nuclear Severe Accident and validate the theoretical models used in calculation codes. In this context, simulated high burnup UO₂ fuels (SIMFUEL) were produced by sintering UO₂ doped with 11 fission products at high temperature. The samples were submitted to  thermal treatments from 400°C up to 1000°C in oxidizing atmosphere. The different phases observed in the samples were identified and finely characterized by Optical Microscopy, Secondary Electron Microscopy coupled with Energy Dispersive X-ray spectroscopy and X-ray Absorption Near Edge Spectroscopy. In particular, Ba and Mo were initially found in two types of phases in the samples:

• Mo was found in metallic precipitates alone or with Ru, Rh and Pd. These precipitates are also observed in Pressurized Water Reactor irradiated fuels in normal operating conditions. 
• Ba was found in oxide precipitates whether as a simple oxide BaO or as (Ba, Sr)(Zr, U, RE)O3. The latter is not observed under PWR normal operating conditions but usually found at higher temperatures. It was thus inferred that the high temperature sintering stage underwent by the SIMFUEL samples enabled the formation of this complex oxide phase.

Detailed post-tests characterizations showed chemical evolution of the phases starting around 900°C. Between 900°C and 1000°C, Mo starts to oxidize to form MoO2. Mo depletion is also observed at temperature as low as 900°C in the Mo-Ru-Rh-Pd precipitates. In the same range of temperature, a reaction between Mo and Ba at the periphery of the oxide precipitates led to partial decomposition of (Ba, Sr)ZrO3 into ZrO2 and BaMoO4. A very good agreement is found between the experimental observations and the mechanisms proposed for Mo and Ba speciation in intermediate oxidizing conditions of a severe accident scenario.

 Thermophysical properties of MOX fuel need to be accurately known to allow reliable predictions from the fuel performance codes regarding the behavior of MOX fuel in reactors. Concerning the stoichiometric MOX U1-yPuyO2, the heat capacity has been computed using the CRG potential [1] through molecular dynamics simulations from 1000 K to the melting point for various Pu contents. Our calculated data highlight a significant effect of the plutonium content only at high temperature (T > ~ 2000 K) where a heat capacity peak around TB = 0.84 Tm (with Tm the melting temperature) is found, associated to the Bredig transition [2], [3] in the literature. We have fitted a behavior law Cp(T,y) on our data, depending on temperature T and Pu content y, valid for the whole range of Pu content and from 300 K to the melting point. The resulting law has then been tested in the fuel performance code GERMINAL [4] by simulating transient tests operated with MOX fuel for fast reactors. We found that our new heat capacity in the solid phase is lower than the reference law currently used in GERMINAL, thus our law yields more penalizing results with respect to the margin to melting. Furthermore, in order to investigate the effect of deviation from stoichiometry on the thermophysical properties of hypostoichiometric MOX U1-yPuyO2-x mixed-oxide fuel, new interactions parameters were optimized in the CRG potential framework. Parameters were fitted to match experimental lattice parameters [5] of U0.70Pu0.30O1.99 (O/M = 1.99) and U0.70Pu0.30O1.97 (O/M = 1.97) through a rigorous empirical procedure combining molecular dynamics and Monte Carlo simulations. Although we fitted our potential only on U0.70Pu0.30O1.99 and U0.70Pu0.30O1.97, we find a very good agreement between the predictions of our potential and the available experimental results as well as the recommendations in the wide range of O/M ratio, Pu content and temperature.

References:

[1]           M. W. D. Cooper, M. J. D. Rushton, et R. W. Grimes, « A many-body potential approach to modelling the thermomechanical properties of actinide oxides », J. Phys. Condens. Matter, vol. 26, no 10, p. 105401, mars 2014, doi: 10.1088/0953-8984/26/10/105401.

[2]           C. Takoukam-Takoundjou, E. Bourasseau, et V. Lachet, « Study of thermodynamic properties of U1-yPuyO2 MOX fuel using classical molecular Monte Carlo simulations », J. Nucl. Mater., p. 152125, mars 2020, doi: 10.1016/j.jnucmat.2020.152125.

[3]           H. Zhang, X. Wang, A. Chremos, et J. F. Douglas, « Superionic UO2: A Model Anharmonic Crystalline Material », p. 57, 2019.

[4]           M. Lainet, B. Michel, J.-C. Dumas, M. Pelletier, et I. Ramière, « GERMINAL, a fuel performance code of the PLEIADES platform to simulate the in-pile behaviour of mixed oxide fuel pins for sodium-cooled fast reactors », J. Nucl. Mater., vol. 516, p. 30‑53, avr. 2019, doi: 10.1016/j.jnucmat.2018.12.030.

[5]           M. Kato, Y. Ikusawa, T. Sunaoshi, A. T. Nelson, et K. J. McClellan, « Thermal expansion measurement of (U,Pu)O 2-x in oxygen partial pressure-controlled atmosphere », J. Nucl. Mater., vol. 469, p. 223‑227, févr. 2016, doi: 10.1016/j.jnucmat.2015.11.048.

Cr-coated zirconium-based cladding materials are being studied and developed within the CEA-Framatome-EDF nuclear fuel joint program, as Enhanced Accident Tolerant Fuel (E-ATF) claddings for Pressurized Water Reactors (PWR) [1] [2]. This paper examines the post-irradiation mechanical behavior of some E-ATF materials. Several first-generation chromium-coated zirconium-based alloys, including recrystallized Zircaloy-4 and Zr-1%Nb alloys, were irradiated up to 2 dpa in the OSIRIS reactor. After irradiation, Expansion Due to Compression (EDC) tests [3] were performed, followed by SEM analysis of the coating surfaces and coating-substrate interfaces.

Tests were performed on Cr-coated Zr-1%Nb materials with coating thicknesses ranging from 2-3 µm up to ~8 µm. Fixed-end EDC tests produce a strain biaxiality close to 0, which is close to that seen in a PWR. The tests indicate the hoop strain at fracture at 350°C. The samples exhibited good ductile behavior, with failure hoop strains in excess of 15%. This high level of strain confirmed the excellent post-irradiation ductility of irradiated Cr-coated cladding tubes.

SEM analyses were used to determine the cracking behavior of the coating as a function of hoop strain, and the integrity of the coating-cladding interface. Comparisons with non-irradiated and previously ion irradiated samples [4] showed negligible effects of irradiation on the coating behavior during mechanical testing.

In summary, Cr-coated materials irradiated up to 2 dpa exhibit good mechanical behavior and coating integrity. SEM examinations coupled with mechanical tests, such as EDC, are a useful tool for characterizing the post-irradiation mechanical behavior of Cr-coated E-ATF claddings under representative in-service loading and thus could be used in the future to test these new materials irradiated at higher doses.

References:

[1] J. Bischoff, et al., “Areva NP’s enhanced accident tolerant fuel developments: focus on Cr-coated M5 cladding”, Nuclear Engineering and Technology 50, (2018), pp. 223-228

[2] J.C. Brachet, et al., “Early studies on Cr-Coated Zircaloy-4 as enhanced accident tolerant nuclear fuel claddings for light water reactors”, JNM 517, (2019), pp. 268-285

[3] M. Rautenberg, et al., “Improvements of PCMI criterion for anticipated operational occurrences”, ENS TopFuel 2018, Prague, Czech Republic

[4] A. Wu, PhD thesis manuscript, (2017), Université Pierre et Marie Curie, France (in French)

M. Bousseksou ^1,2, D. Gosset², G. Gutierrez¹, Y. Pipon³, N. Moncoffre³

¹CEA, DEN/SRMP JANNUS, Gif-sur-Yvette, France

²CEA, DEN/SRMA LA2M, Gif-sur-Yvette, France

³CNRS, IP2I, Lyon, France

Boron carbide is used as a neutron absorber for reactors control and installations protection. In the reactor, the absorption reactions (¹⁰B (n, α) ⁷Li and ¹⁰B (n, 2a) ³H) lead to the formation of helium, lithium and tritium atoms. The slowing down and stopping of these atoms together with those displaced from capture-less fast neutron interactions lead to several hundred atomic displacements per effective neutron capture, with resulting dpa (displacements per B₄C atom) rates similar to those observed in the fuel claddings. The behaviour of helium is well understood. It is mostly retained, up to high temperature, in the form of highly pressurized bubbles, leading to swelling and cracking of the material [1]. Conversely, very few results are available concerning lithium behaviour and the associated diffusion mechanisms [2,3]. Only a low apparent effect with a grain boundaries cohesion loss was observed [3].

This work aims to analyse the lithium behaviour in B₄C and first to determine the lithium coefficient diffusion. High density B₄C samples were obtained by Spark Plasma Sintering. The lithium atoms were introduced in the matrix by using ion implantation. Annealing treatments were performed to activate the lithium diffusion. To obtain the Li concentration profiles as a function of depth, the samples were analysed by Secondary Ionization Mass Spectroscopy before and after each treatment. The diffusion coefficients are deduced from their evolution. The damage build up was investigated by Raman spectroscopy. The first results will be presented.

[1]      V. Motte, “Behavior of helium in implanted boron carbide ”, PhD Thesis, Lyon-1 University (2017).

[2]      X. Deschanels et al., JNM 265(1999) 321-324.

[3]      P.D. Kervalishvili, et al., At. Energiya 57-7 (1984) 52-53.

Modelling approach for depth profiling of radiological contamination in concrete, using CT-scan images

Lowie Brabants^1*, Guillaume Lutter², Jan Paepen², Bram Vandoren³, Wouter Schroeyers¹ 

1 Hasselt University, NuTeC, CMK, Nuclear Technology - Faculty of Engineering Technology, Agoralaan building H, B-3590 Diepenbeek, Belgium

2 EC-JRC-IRMM, European Commission, Joint Research Centre, Institute for Reference Materials and Measurements, Retieseweg 111, Geel 2440, Belgium

3 Hasselt University, CERG, Faculty of Engineering Technology, Agoralaan building H, B-3590 Diepenbeek,
  Belgium

 * Main Author, E-mail: lowie.brabants@uhasselt.be

 Keywords:

Concrete, non-destructive testing, CT-scan, depth profiling

 GRAPHICAL ABSTRACT

ABSTRACT

Concrete is the main fraction of waste coming from the decommissioning of nuclear power plants (NPPs). Assigning the correct waste category for this concrete is essential in reducing costs when decommissioning NPPs. For this purpose, a decontamination plan is created which aims at decontaminating the structure by removing several mm-cm of concrete. This plan is based on a radiological analysis which maps areas with elevated radionuclide activities, and which assesses the depth of the contamination. 

Depth profiling is currently performed using a combination of measurement techniques. Destructive methods, such as core drilling, can be time consuming and require extensive sample preparation. Non-destructive measurement techniques, such as the relative linear attenuation model or multi-collimator methods, experience difficulties when profiling depth in concrete samples as the concrete is inhomogeneous, consisting of aggregates, voids and mortar. Furthermore, selection of the best depth profiling method is cumbersome as there is no reference concrete source or model available to benchmark the different techniques. 

This study aims at developing a Monte Carlo model which will incorporate the complex structure of concrete, as well as serve as a model for benchmarking or developing different depth profiling techniques. In a first step, concrete samples, with a mixing design similar to concrete used in NPPs, were created and imaged using a CT-scanner. The resulting images give insight in the complex density and chemical composition of the concrete samples. In the next step, these samples were contaminated trough contact with a solution containing key radionuclides, such as ¹³⁷Cs, encountered during nuclear decommissioning. The resulting contaminated samples will then be destructively analyzed in a layer by layer approach to obtain a contamination depth profile. The obtained contamination profile in combination with the CT-scan images will then be used to construct the Monte Carlo model.

Zirconium-based claddings with an outer chromium coating resistant to corrosion are studied and developed as an evolutionary Enhanced Accident Tolerant Fuel (E-ATF) concept for light water reactors [1]. However, in hypothetical LOss-of-Coolant-Accident (LOCA) conditions, following clad ballooning and burst, the outer coating does not allow to protect the inner surface of the cladding from High Temperature (HT) steam oxidation and associated secondary hydriding (i.e., massive localized hydrogen uptake which may induce potential clad failure upon the water quenching at the end of the transient).

To address this issue, DLIMO-CVD (Direct Liquid Injection of Metal-Organic precursors - Chemical Vapor Deposition) CrxCy coatings have been developed and successfully deposited onto the inner surface of Zr-based cladding tube prototypes. Then, preliminary two-sided oxidation tests have shown that such inner coating is able to increase the resistance to oxidation at HT of the inner clad surface [2].

The present study aimed at performing new steam oxidation tests at 1200°C on Zr-based clad prototypes with a CrxCy inner coating, in conditions more representative of LOCA, after a first internal pressure-induced burst step. Additionally, complementary two-sided steam oxidation tests have been carried out up to 1h at 1200°C, on short inner and/or outer-coated clad segments. Finally, Post-Quench (PQ) Ring Compression Tests, fractographic analysis and deep metallurgical investigations including neutron-tomography have been performed to get more insights into the PQ behavior of the inner-coated clad.

Among other results, it is shown that the inner CrxCy coating makes it possible to reduce significantly the oxidation and the associated secondary hydriding of the clad inner surface, after ballooning and burst. Consequently, after at least 600s under steam at 1200°C (i.e., for an ECR“Baker-Just“ value of at least 30%), it is shown that the reference uncoated clad failed upon final water quenching while the inner-coated prototype kept its integrity.

References:

[1] J.C. Brachet, et al., JNM 517 (2019) 268-285

[2] A. Michau, et al., SCT (2018), DOI:10.1016/J.SURFCOAT.2018.05.088

Oxidation of High Density Fuels Investigated with Hard X-ray Photoelectron Spectroscopy

J. Buckley¹, B. F. Spencer², R. N. Worth¹, R. W. Harrison¹, W. R. Flavell³, and

 T. J. Abram¹

¹Department of Mechanical, Aerospace and Civil Engineering, The University of Manchester, Oxford Road, Manchester, United Kingdom

²Department of Materials, The University of Manchester, Oxford Road, Manchester, United Kingdom

³Department of Physics and Astronomy, The University of Manchester, Oxford Road, Manchester, United Kingdom

High uranium density fuels, such as U₃Si₂ and UN, offer significant neutronic and safety related benefits over UO₂, when considered as a replacement water cooled reactor fuel. However, the oxidation behaviour of these materials requires further understanding and improvement before wide-scale deployment can be considered. The initial oxidation of high density fuels has been investigated using Hard X-ray Photoelectron Spectroscopy (HAXPES). This technique probes the electronic structure of the material beyond the immediate surface layer, as is limited by conventional XPS. The novel application of this technique has been used to non-destructively investigate the oxide layers formed when materials are exposed to different conditions, at the surface and progressing into the bulk material. This allows for an oxidised layer thickness to be extracted, and a greater understanding is gained into the formation of intermediate products. This work therefore demonstrates the capability to probe the bulk properties of air-sensitive radioactive materials without the requirement to etch or sputter the surface. This reduces the risk of contamination, enabling the uptake of this technique in facilities that would not typically accept such materials.

The eutectic alloy of lithium-lead (PbLi) is one of the proposed compounds for tritium breeding in the future fusion power plants. The contact of structural materials such as EUROFER or others RAFM steels with the PbLi causes a corrosion layer that can have other important consequences on the safety and the right functioning of the fusion reactor. Over the last years, studies on the behavior of structural materials subjected to corrosion have put in evidence the need of new facilities with renewed experimental conditions.

In the new Spanish Liquid Metal Laboratory (LML), activities are focused on the experimental validation of liquid metals-based technologies. The study of open questions concerning the effect of PbLi on structural and functional materials is one of its objectives. The aim is to observe not only the corrosion effect on the structural material itself but also to study the new materials proposed as coating. To address these issues, two new facilities have been developed to work in parallel under static and dynamic conditions, respectively.  The first one is called COES and the second one, CICLO. 

The Corrosion Experiment in Static condition (COES) features a fully internally coated vessel that will prevent corrosion of the test section during long-term experiments, providing better results with this first static test.

The Ciemat Corrosio Loop (CiCLo) has been designed to be able to work in the strongest breeder blanket conditions: up to 550ºC and at 1m/s in its tubular test section. These are without a doubt the most extreme conditions that a dynamic loop installation could provide today.

This work shows a detailed presentation of all the possibilities offered by the first corrosion facility in Spain and a description of their installation and commissioning.

In Fast Breeder Reactors, the large thermal gradient within the (U,Pu)O_2±x fuel induces oxygen and metal mass transport along the pellet radius. The aim of the present work is to build a mobility database for plutonium diffusion in (U,Pu)O_2±x using the DICTRA software. This mobility database is based on the well-established thermodynamic description of the mixed oxide in the (U,Pu)O_2±x [1] and does extend  the oxygen diffusion database previously published in [2,3]. The mobility parameters for Pu³⁺ and Pu⁴⁺ are adjusted using carefully selected experimental information and data obtained with the help of the so-called cBΩ model for plutonia [4]. An overall reasonnable agreement is obtained with the experimental data. The model allows the description of plutonium diffusion in the UO_2±x-PuO_2-x region for any plutonium content, oxygen stoichiometry and temperature thanks to the change in the charge of cations (Pu³⁺,Pu⁴⁺,U³⁺,U⁴⁺,U⁵⁺) and oxygen defects (vacancy and interstitial) with composition and temperature. The approach and the results will be presented.

[1] C. Guéneau et al, J. Nucl. Mater. 419 (2011) 145

[2] E. Moore et al, J. Sol. State Chem. 203 (2013) 145.

[3] E. Moore et al, J. Nucl. Mater. 485 (2017) 216

[4] P. Varotsos, K. Alexopoulos, Phys. Rev. B24 (1981) 904.

Liquid media based on Pb or Pb-Li are planned for cooling of the future fission reactors. Such molten metals environment represents very aggressive conditions for the used structural materials. Owing to their unique self-passivation principles, ferritic 9-16%Cr ODS-based steels represent promising candidate materials, capable to withstand the severe conditions of molten Pb. However, in the case of molten Pb-Li environment, this strategy cannot be used due to the higher affinity of lithium to oxygen when compared with chromium. The alternative approach for Pb-Li may reside in a deposition of thick protective coatings made of low-solubility hard pure metals, such as W or Mo.

Radio-frequency inductively-coupled plasma spray (RF-ICP) is a process of coatings formation via melting of powder feedstock and its acceleration toward the substrate materials. The technology is capable of deposition of ceramic materials, and, given its controlled protective atmosphere, also of (sensitive) metals. Importantly, the technology is readily up-scalable, allowing covering of large areas using industrial high power systems. Given the immense plasma heat input, it was originally impossible to deposit the coatings onto materials such as stainless steel. To overcome this limitation, an internally water-cooled holder was designed, allowing to use materials more relevant for the targeted application.

In this study, thick molybdenum coatings deposited on AISI 304 coupons are presented. Three different RF-ICP torch power levels were tested and the respective microstructures, chemical compositions and the interface quality of the coatings are discussed with respect to the intended application.

Tungsten is a candidate material for the plasma facing component of the future power plants by magnetic fusion. The tungsten divertor of ITER and the wall of DEMO will be in contact with the hydrogen isotope plasma and will be bombarded by energetic ions and neutrons. To understand the formation of defects and the evolution of the microstructure, dedicated experiments and their multiscale modelling counterparts are necessary. In this work, (i) a systematic experimental study (using Transmission Electron Microscopy) of samples implanted with 1.2 MeV W ions and annealed at different temperatures up to 1800 K along with (ii) an object kinetic Monte Carlo modelling coupled to an optimization solver (OPENTURNs) to obtain the diffusion of vacancy defect as function of their size (which is poorly known) have been done. The resulting model gives good results to reproduce other experiments. It is also easily extendable to neutron irradiation.

Additional novelties of this work are, firstly, that we built the source term of the model by picking in a very large database of Molecular Dynamics cascades of energy varying from 0.25 to 300 keV in W and by combining with the decomposition in subcascades of the high energy ion impacts. Secondly, the model accounts for the role of main commercial tungsten impurities like C, O, P, Fe, Mo, He, H, N, Co. Density Functional Theory calculations of the binding energies of the mono vacancies, the mono self interstitial atoms and the small dislocation loops, fed our model for the interaction of defects with impurities, namely the trapping or the releasing of the point defects as a function of the defect size and temperature. Our simulations indicate that the impurities have an important role in trapping SIA dislocation loops up to high temperature, 1200 K, however they should not trap nanocavities.

During reactor operation, the extreme temperature and radiation conditions lead to deep modifications in the microstructure of the nuclear fuel. The high neutron flux experienced by the rim of the fuel results into the development of the High Burnup Structure (HBS). In this region, the original 10-15µm grains undergo polygonization leading to a final microstructure composed of 100-300nm grains surrounding micrometric pores. During repository, the HBS will constitute the interface of the fuel with the external environment in case of failure of the confinement. For this reason, it is fundamental to deeply understand its degradation behaviour when subjected to abnormal storage conditions.

While oxidation mechanisms of UO₂ are widely reported in the literature, the effect of the grain size on the oxidation kinetic has not been investigated yet. Studies on powders highlighted the existence of a threshold size under which cracking of the particles does not occur, and oxidation is hence limited to U₃O₇ instead of proceeding to U₃O₈. However, no similar work was performed on densified UO₂ pellet.

In the present study, UO₂ disks were densified from nanopowders using Spark Plasma Sintering (SPS), which allowed limiting coarsening during sintering. In this way, a set of identical samples (in terms of density and O/M ratio) differing only by the grain size (3µm, 500nm, 200nm, such as shown in the Figure) was prepared. Their oxidation behaviour was studied by in-situ High Temperature Synchrotron Radiation X-ray Diffraction (XRD) and X-ray Absorption Near Edge Structure (XANES) at the Rossendorf Beamline (ESRF, Grenoble). Thanks to this unique combination, we were able to assess the structure (space group and phase repartition) and the O/M ratio of different grain size UO₂ as a function of the temperature.  Contrarily to the other two samples, no U₃O₈ could be detected in the sample of 200nm of grain size.

NDT study of chromium rich bcc α' phase precipitation in Fe20Cr ODS alloy

J. Degmova, V. Krsjak, S. Sojak, M. Kotvas, J. Dekan , M. Petriska, P. Mikula

Institute of Nuclear and Physical Engineering, Slovak University of Technology, Ilkovicova 3, 81219 Bratislava, Slovakia

Abstract

There is still some concern regarding the ferritic, dual-phase and martensitic steels with Cr content higher than 12 wt. % known as “475 °C embrittlement”. This phenomenon is attributed to phase separation of the α-(Fe, Cr) ferrite phase into Fe-rich ferrite (α) and the Cr-rich ferrite (α') phases in the miscibility gap which exists below 500 °C in the Fe–Cr binary system. 

In the present paper, we focused on the determination of α' precipitation in PM2000 (Fe20Cr) steel by various non-destructive techniques as magnetic Barkhausen noise (MBN), positron annihilation spectroscopy (PAS) and Mossbauer spectroscopy (MS). 

These non-destructive techniques were used in a complementary manner to determine α' phase in the PM2000 samples isothermally annealed at 475 °C up to 1000 h. The same annealing was performed on T91 (Fe8Cr) steel in order to compare the differences in the obtained results with the alloys where α' development is not expected. 

Key words: 475 °C embrittlement, magnetic Barkhausen noise, positron annihilation spectroscopy, Mossbauer spectroscopy

Thermodynamic assessment of the LiF-NaF-KF-CrF₂-CrF₃ system

T. Dumaire^a, R. Hania^b, R.J.M. Konings^c, A.L. Smith^a

^a Delft University of Technology (Delft, Netherlands)

^b NRG (Petten, Netherlands)

^c European Commission, Joint Research Centre (Karlsruhe, Germany)

Keywords: FLiNaK, chromium fluoride, corrosion, calorimetry, CALPHAD 

Fluoride salts are considered a promising heat transfer medium and coolant in the primary and secondary loops of the Generation IV Molten Salt Reactors (MSRs) [1]. A major concern for the operation of these reactors, is the degradation of the structural materials caused by the corrosive properties of the fluoride salt at high temperatures. The redox potential of the fuel salt, controlled by the UF₄/UF₃ ratio, determines the rate of corrosion of the structural material. During irradiation, free fluorine is formed which reacts with UF₃, hence increasing the UF₄/UF₃ ratio and redox potential of the salt. This results in an increase in the corrosion rate of the structural material. Chromium is the least stable element in the Ni-based alloys (especially Hastelloy-N) envisaged in the fluoride-fuelled MSR, with oxidation reactions such as Cr(alloy)+ 2UF₄(salt) = CrF₂(salt)+2UF₃(salt) [2].  The preferential removal of chromium leads to the formation of discrete voids in the Ni-based alloy [3], and dissolution of this element in the fluoride salt.

The understanding of the corrosion mechanisms and the effect of the corrosion products on the basic properties of the salt (melting point, heat capacity etc.) is fundamental for the safety and durability of this technology. This work focuses on the thermodynamic assessment of the CrF₂-CrF₃ system and the ternary and quaternary systems with alkali fluorides (FLiNaK), coupling experimental phase diagram investigations using calorimetry and thermodynamic modelling assessments using the CALPHAD (Computer Coupling of Phase Diagrams and Thermochemistry) method and the quasi-chemical formalism in the quadruplet approximation.

Bibliography

Within the PLEIADES simulation platform co-developed by CEA, EDF and FRAMATOME, Germinal is the fuel performance code devoted to the simulation of the in-pile behaviour of mixed oxide fuel pins for Sodium-cooled Fast Reactors. The feedback on such fuel elements, based on experimental observations mainly acquired during Phénix french reactor operation, shows that a layer of volatile fission products, called JOG for “Joint Oxyde-Gaine”, is formed between the fuel pellet and the clad for a burn-up of 6 to 8 %FIMA. At higher burnup, the volatile fission products (VFP) compounds can react with the cladding material components (Fe, Ni, Cr), corroding the inner wall of the clad.

Thus, it is important to improve our chemical modeling of the uranium and plutonium mixed oxide fuel behavior under typical irradiation conditions. For that purpose, the thermochemical code ANGE and the OpenCalphad (OC) software have been integrated into the Germinal fuel performance code.

Simulations with GERMINAL have been performed on French reactor Phénix fuel pins irradiated to different burnup both with ANGE + TBASE using a description based on that proposed by Lindemer & Brynestad, and OpenCalphad + TAF-ID using the description of Guéneau et al. for the modeling of the fuel with its fission products in solution.

The results in terms of JOG thickness and of fuel pellet chemical composition are compared to post-irradiation observations. This analysis shows that the simulations are strongly dependent on the hypothesis considered for the volatile fission products release rate as well as on the oxygen content considered in the calculations, for which some examples are illustrated in this study.

The material under study is a nuclear fuel made up of plutonium (Pu) and uranium oxides. The objective of this work is to generate a representative elementary volume (REV) of its microstructure, characterized by the spatial distribution of the Pu content, and to develop on it a reduced micromechanical model using the Nonuniform Transformation Field Analysis (NTFA) method.

The texture of the material is recognized by maps of Pu content, obtained by electron microprobe. In the past, the material was modeled as a three-phase medium composed of two inclusion phases dispersed in a matrix, with uniform Pu content in the three phases. A new modeling, based on a combination of random sets and random functions, improves the generated microstructures by allowing the Pu content to vary in each phase. The histogram and the spatial covariance of Pu content, which are two fundamental statistical characteristics of the variability, are calculated from the maps of Pu content. Special care is taken to the measurement noise associated to the electron microprobe. The variability is then simulated with respect to these calculated histogram and spatial covariance.

From a mechanical point of view, the nuclear fuel is a material with non-linear ageing viscoelastic behavior and free swelling, which depends on the local plutonium content. Mechanical calculations, using a Fast Fourier Transform resolution method, were performed on both microstructures (with and without Pu variability in each phase), and the results show that the effect of variability is important. On the other hand, the NTFA method, previously used for the same material~\cite{largenton2014}, is extended to take into account the Pu content variability, and optimized by linearizing the non-linear behavior using the tangent second-order approximation (NTFA-TSO~\cite{michel16}). The predictions of the new reduced model are in good agreement with full-field simulations with a considerable gain in computational time.

This work was supported by the I3P research institute (CEA-EDF-FRAMATOME).

Since the 1950s, nuclear fusion reactors have been heralded as the energy source of the future. With a promise of clean reliable energy, many people believe that it is one of the answers to the current energy crisis. In order for fusion energy to become economically viable, plasma facing components must be able to withstand the harsh reactor environment and be reasonably costed and simple to manufacture and repair.

Previous work performed in collaboration between Oxford and UKAEA-CCFE has developed novel “sculpted” substrates which allow coatings of up to 4 mm thickness to be manufactured repeatedly using vacuum plasma spraying (VPS) on a stainless steel substrate. However there is a lack of understanding on how the microstructure controls the key thermo-mechanical properties of the coatings. In literature, sprayed W coatings rarely exceed 1 mm, however more comprehensive characterisation means that these coatings are better understood.

The present work aims to establish a link between these two approaches: combining ultra-thick (>2 mm) tungsten coatings making using of a patterned substrate; and utilisation of advanced characterisation and mechanical testing methods to study properties and relate them to processing conditions. Emphasis is given to minimisation and management of thermally induced residual strains by inducing controlled micro-cracking via a micro-millimetre scale patterned substrate, as a target coating thickness of 5 mm is realised. A combination of process details, microstructure and mechanical assessment is presented, and the system is shown to have promise for technological application in the civil nuclear power sector.

Zirconium alloys are used in nuclear industry as cladding materials due to their low neutron absorption and good mechanical performance at high temperature. During service, the corrosion of the cladding generates hydrogen that at the operation temperature tends to diffuse into the metal. Hydrogen, in solid solution and in its precipitated form, changes the cladding mechanical properties. The mechanical properties are relevant in the context of safety of spent nuclear fuel handling and transport operations.

When hydrogen is in solid solution it can influence the dislocation mobility by shielding the interaction of dislocations with other types of defects, like grain boundaries and second phases. This phenomena, known as Hydrogen Enhanced Localized Plasticity (HELP), has been extensively studied in FCC and BCC metals, but a complete understanding and description of the phenomena in Zr based alloys and HCP metals in general is lacking.

Nanoindentation measurements, at  high-temperature, have been performed on hidrogenated and non-hydrogenated Zircaloy-4 samples. The experiments aims at verifing the inluence of hydrogen under solid solibity and  also at defining the boundary conditions of hydrogen concentration, deformation rate, and temperatures necessary to observe the HELP effect over other competitive H-embrittelment mechanisms active in zirconium alloys. Results and and possible implications will be presented at this conference. 

In pressurized water reactor (PWR) conditions, helium is produced from transmutation of nickel in austenitic stainless steels, at levels of around 10 appm/dpa. Such high levels of helium production can impact the microstructure, in particular by forming helium bubbles. The precise effect of helium in these conditions is still badly unknown. To better understand the role of helium, a two-phase study was performed.
The first part of the study consisted in using molecular dynamics to parameterize a "variable-gap model" [1] for helium bubbles in FCC nickel. Nickel was used as a model material for austenitic stainless steels. This model predicts bubbles' binding energies, which is an important quantity that describes their stability.
In a second step, a variable-gap model was implemented in a CD code and it was used to simulate the evolution of defects in nickel under PWR irradiation conditions with focus on the evolution and stability of bubbles.
[1] T. Jourdan and J. P. Crocombette, J. Nucl. Mater. 418, 98 (2011)

Uranium nitrides are an attractive alternative fuel form to oxides, particularly for fast reactors, due to their higher thermal conductivity, higher fissile element density whilst still exhibiting a high melting point and good radiation tolerance. Furthermore, nitrides accommodate a broad range of non-stoichiometry. Although these properties should allow the fuel to operate at higher burnups they will change because of defect formation.   Consequently, we have used atomic scale simulations to predict how specific defect configurations influence properties.  Initially we use a quantum mechanical approach based on the DFT approximation. Using this we report predictions of how defects accommodate non-stoichiometry and their influence on structural and mechanical properties.  Surface structure predictions are also analyzed. Using these predictions we can employ new and existing empirical potential models to predict temperature dependent properties. This allows us to investigate diffusion rates that require a point defect model to be used in higher level codes such as cluster dynamics. Thus, the role for DFT calculations is also to provide an input to these higher-level models and to understand the behaviour of defects more fully in uranium nitride fuel.

Composite fuels such as UN-UO₂ have been proposed to overcome the lower oxidation resistance of the UN fuel. An innovative UN-UO₂ accident tolerant fuel has recently been fabricated and studied by our group: UN microspheres embedded in UO₂ matrix. In the present study, detailed thermogravimetric investigations (TG/DSC) of high density UN-UO₂ composite fuels (91-97 %TD) are reported. Three samples of each specimen were heated at 5 °C up to 700 °C in air to understand the material parameters which govern the oxidation reactions of these innovative UN-UO₂ fuels. This work also describes the influence of UN addition (10-50 wt%), as well as the amount of phases present in the sintered samples, on the oxidation reactions. The results show that the UO₂ phase significantly increased the oxidation onset temperature (OOT) of the fuel from 260(3) °C (pure UN pellet) to 320(4) °C (with 10 wt% of UN). The composites with 30 wt% and 50 wt% of UN presented similar OOT (~285 °C), which is close to pure UO₂ pellet (~300 °C). Furthermore, the total weight variations (%) and the maximum reaction temperatures (MRTs) are reported and discussed in this study.

Introduction: Heavy liquid metals (HLM), lead and lead-bismuth (LBE), are excellent cooling media for nuclear systems of next generation but most structure materials suffer from general corrosion in contact with them. Ferritic-martensitic steels (FM), containing 14-8 wt.% Cr, had been considered appropriate candidates to build cores of Gen IV nuclear systems owing good resistance to neutron radiation and oxidation. Performance of FMs had been widely studied^1-8; research recently focused to 400 – 550°C and 10^-8–10^-6 wt.% O, oxygen being corrosion inhibitor⁹. The 9Cr FMs behaved fair and formed double- or triple-oxide scales. Most of data exist for LBE at 490-550 °C, giving about 50 μm oxide in 3000 hours in flowing LBE^4-8,10,11; at 550 °C the scale failed and solution-based attacks (SBA) occured¹². Data of the FM oxidation in liquid lead are scarce, giving about 50 μm in 5000 hours in flowing lead^2,9,13. Moreover, the FMs had been found sensitive to environmentally assisted cracking (EAC/LME); since no real embrittlement in HLM was discovered ^14-16, old term LME changed to EAC/LME. Despite prerequisite of stress&strain levels far beyond design¹⁵, EAC/LME has become an issue for the HLM-cooled systems. Consequently, 15-15Ti stainless steel was selected for core of the first Lead-Cooled Fast Reactors demonstrator ALFRED¹⁴. But the 15-15Ti, having less corrosion and no EAC/LME, is not optimal solution, as it suffers from SBA in HLM.

Methods: Grade91 and 15-15Ti cylindrical samples exposed to liquid lead (480°C, 10^-7 wt.% O, 2 m/s) for 8000 hours, next post-test microscopy investigation.

Results: after 8000h exposure

• P91: <20 μm triple-layer scale; without SBA.
• 15-15Ti: <10 μm triple-layer scale; SBA indications; outer oxide exfoliation.

Discussion: Is the present choice of 15-15Ti, exfoliating thin scales and suffering from SBA developing into pits, better than the Grade91, which grows thick oxide scales worsening heat transfer?

References: 1. Fazio2001, 2.Glasbrenner2001, 3.Martín2004, 4.Muller2004, 5.Kurata2010, 6.Konys2015, 7.Schroer2014, 8.Tsisar2017, 9.Zhang2009, 10.Aiello2004, 11.Weisenburger2011, 12.Shi2019, 13.Gessi2008, 14.Glickman2011, 15.Hojna2020, 16.Serre2018, 17.Alemberti2014