Experiments comparing the embrittling effects of high-energy (10-MeV) electrons and test reactor fast neutrons have been performed to elucidate the role of primary damage state on reactor pressure vessel embrittlement. Electrons produce displacement damage primarily by low-energy atomic recoils, while fast neutrons produce displacements from considerably higher energy recoils. Comparison of changes resulting from neutron irradiation, in which nascent point-defect clusters can form in dense cascades, with electron irradiation, where cascade formation is minimized, can provide insight into the role that the in-cascade point-defect clusters have on embrittlement mechanisms. Yield strength changes induced by 10-MeV electrons or test reactor neutron irradiations of unalloyed iron and an Fe-0.9 wt.% Cu-1.0 wt.% Mn alloy were examined up to 1.5 x 10-2 dpa. The small alloying additions substantially enhanced embrittlement in both the electron- and neutron-irradiated samples relative to unalloyed iron. Similar embrittlement trends with increasing radiation damage were observed for electrons and neutrons in both the ternary and unalloyed iron. Preliminary results of small-angle X-ray scattering measurements of embrittled microstructures are also reported.

Comparative experiments using high energy (10 MeV) electrons and test reactor neutrons have been undertaken to understand the role that primary damage state has on hardening (embrittlement) induced by irradiation at 300 degreesC. Electrons produce displacement damage primarily by low energy atomic recoils, while fast neutrons produce displacements from considerably higher energy recoils. Comparison of changes resulting from neutron irradiation, in which nascent point defect clusters can form in dense cascades, with electron irradiation, where cascade formation is minimized, can provide insight into the role that the in-cascade point defect clusters have on the mechanisms of embrittlement. Tensile property changes induced by 10 MeV electrons or test reactor neutron irradiations of unalloyed iron and an Fe-0.9 wt.% Cu-1.0 wt.% Mn alloy were examined in the damage range of 9.0x10(-5) dpa to 1.5x10(-2) dpa. The results to date showed the ternary alloy experienced substantially greater embrittlement in both the electron and neutron irradiated samples relative to unalloyed iron. Surprisingly, despite their disparate nature of defect production, similar embrittlement trends with increasing radiation damage were observed for electrons and neutrons in both the ternary and unalloyed iron.

We analyze the post-yield true-stress vs. true-strain flow behavior of neutron and electron irradiated reactor pressure vessel steels, A212B and A350, and binary alloys, Fe-0.28 Cu and Fe-0.74 Ni. The flow curves suggest that neutron-irradiation hardening has the same effect as strain hardening for all the materials examined. The post-yield flow curves, obtained after electron-irradiation hardening to yield strength levels similar to those achieved by neutron irradiation, behave differently.

Motivated by the recent interest in gamma ray embrittlement of nuclear reactor pressure vessels (RPVs), calcultions were performed to evaluate aspects of defect production by gammas in iron and steel, in addition to determining displacement damage cross-sections, the atomic recoil energy dependence of gamma-induced defect production was described by integral recoil damage spectra, W(T), and their associated median recoil damage energies, T-1/2. These latter characterizations, should be particularly useful in evaluating the contribution of gamma ray generated defects to microstructural changes causing radiation embrittlement. The results for monoenergetic gammas, as well as for gamma rays with a spectrum of energies characteristic of a RPV, reveal T-1/2 values of < 100 eV, about three orders of magnitude smaller than for fast-neutrons, the radiation of primary concern in previous embrittlement studies. The relative contributions of various gamma interactions to defect production, as well as the role of light alloy element-induced secondary displacement mechanisms, are also considered.

ＴＥＭＳＴＵＤＩＥＳＦＯＲＤＩＧＭＩＮＫｒＩＯＮＩＲＲＡＤＩＡＴＥＤＡｕ－ＣｕＢＩＬＡＹＥＲＳ￥ＧａｏＹｕｚｕｎ；ＡｌｅｘａｎｄｅｒＤＥ；ＲｅｈｎＬＥ（ＧｅｎｅｒａｌＲｅｓｅａｒｃｈｉｎｓｔｉｔｕｔｅｆｏｒＮｏｎｆｅｒｒｏｕｓＭｅｔａｌｓ，Ｂｅｉｊｉ...

We demonstrate that high-energy electron irradiation can be used to study gamma-ray embrittlement of reactor pressure vessels. Observed changes in tensile properties, in ferritic alloys electron irradiated at low temperature (⩽ 50°C), correlate well with neutron irradiation data when compared simply on a displacement per atom basis. The similarity in embrittlement behavior between gamma-rays (electrons) and neutrons can be explained in terms of a reaction-rate theory model of defect production and clustering. The present results, obtained under well-controlled experimental conditions, provide strong support for a previous analysis implicating gamma-ray damage as the source of the “accelerated” embrittlement of the High Flux Isotope Reactor (HFIR) pressure vessel.

Recent work addressing the enhanced interdiffusion rates typically observed during ion-beam mixing at elevated temperatures is reviewed. As discussed previously, the expected increase in ion-beam mixing rates due to “radiation-enhanced diffusion” (RED), i.e. the free migration of isolated vacancy and interstitial defects, is well documented in single-crystal specimens in the range of 0.4 to 0.6 of the absolute melting temperature. In contrast, the increase often observed at somewhat lower temperatures during ion-beam mixing of polycrystalline specimens is not well understood. However, sufficient evidence is available to show that this increase reflects intracascade enhancement of a thermally-activated process that also occurs in the absence of irradiation. Recent evidence is presented which suggests that this process is Diffusion-Induced Grain-Boundary Migration (DIGM). An important complementary conclusion is that because ion-beam mixing in single-crystal specimens exhibits no significant temperature dependence below that of RED, models that invoke only irradiation-specific phenomena, e.g., cascade-overlap, thermal-spikes, or liquid-diffusion, and hence which predict no difference in mixing behavior between single- or poly-crystalline specimens, cannot account for the existing results.

The neutron irradiation-induced mechanical property change is a major challenge for nuclear materials. For a Pressurized Water Reactor (PWR), neutron irradiation embrittlement of the reactor pressure vessel is one of the most important issues that determine the whole nuclear power plant lifetime. The present work focuses thus on the uncertainty assessment for the neutron irradiation displacement damage of a PWR vessel. The uncertainties of neutron flux, nuclear data, and atomic displacement models are propagated to Displacement per Atom (DPA) rates with a special emphasis on the importance of correlation matrices. In addition to the uncertainty propagated from the threshold displacement energy, the total uncertainty of damage energy rate is about 10% (at 1σ), whereas the uncertainty is largely underestimated if the correlations of damage cross section and neutron flux are not considered. Apart from the complete uncertainty propagation shown in this paper, combining the damage energy rate uncertainty and the displacement threshold energy uncertainty is recommended for estimating the calculated DPA rate uncertainty. Besides the uncertainty, the biases induced by the model defects are preliminarily estimated.

The present work investigates the attenuation of neutron and photon-induced irradiation damage in a Pressurized Water Reactor (PWR)-like mock-up Reactor Pressure Vessel (RPV) using Tripoli-4® Monte Carlo simulations with a particular emphasis on the presence of a thick stainless steel heavy reflector between the core and RPV. Results show that the photon-induced damage is well described by the exponential law. The neutron-induced damage attenuates quicker than the exponential form near the outer surface of RPV. Nevertheless, an exponential form can represent the attenuation of neutron-induced damage within 5% discrepancy for penetration < 17 cm in a typical 22 cm thick RPV and within 20% in the outermost 3 cm. The exponential form with an additional negative term fits the attenuation of neutron-induced damage in RPV as the negative term considers the “leakage” of neutron near the outer surface. It is observed that the presence of a Gen III-like representative heavy reflector reduces the estimated neutron-induced damage by a factor of 2. On the other hand, this paper verifies that the neutron flux with energies above 0.5 MeV (ϕ>0.5) is more representative of the displacement damage attenuation than the flux above 0, 0.1, and 1 MeV in RPVs. The damage rate is approximatively equal to Kϕ>0.5 with K=9.5×10-22DPA/n∙cm-2 in PWR RPVs.

High-angle X-ray diffraction was used to determine cascade mixing efficiencies, Dt/dpa, in Nb/V superlattices irradiated to small doses (≤ 0.26 dpa). Samples were neutron irradiated in the core of the High Flux Beam Reactor at Brookhaven National Laboratory and ion irradiated with 1.5 MeV Ne. No significant differences were observed in cascade mixing efficiencies between the two types of irradiation. Values of Dt/dpa were observed to vary with the modulation wavelength of the superlattice, with thicker wavelength samples yielding values approaching that determined from ion mixing thick Nb/V bilayers ( Dt/ dpa ∼ 110 A 2 / dpa ) . The decrease in mixing efficiency observed at lower wavelengths may be related to an observed structural transition in which the interfaces of the superlattice become coherent.

Grain growth commonly observed during heavy-ion irradiation of initially fine-grained (less-than-or-equal-to 100 angstrom diameter) thin films is modeled as a thermal-spike phenomenon in which temperature spikes caused by ions and recoils induce atomic jumps across grain boundaries, promoting boundary migration. In elemental and homogeneous alloy systems, in which grain growth is driven solely by the reduction of boundary surface area, the model predicts that the ion-induced grain-boundary mobility is linearly proportional to the quantity, F(D)2/DELTAH(coh)3, in which F(D) is the ion and recoil energy deposited in elastic collisions and DELTAH(coh) is the cohesive energy of the target. The model was evaluated with respect to data from two previously published ion-induced grain-growth experiments on elemental and coevaporated alloy films. The results were consistent with the thermal-spike model. Combining analytical results of the model with the experimental data it was possible to determine the value of the proportionality constant beta relating the cohesive energy to the activation energy Q for grain growth (Q = -betaDELTAH(coh)). The value of beta for the coevaporated and elemental films, respectively, was 0.07 and 0.15, which is less than or about equal to the value previously determined for the thermal-spike treatment of ion beam mixing (beta(IM) = 0.14). The smaller value of beta determined for the coevaporated films is consistent with the idea that atom migration across grain boundaries is easier than migration within the lattice. The thermal-spike treatment was also applied to ion-induced grain growth in multilayer films. The presence of concentration gradients in these systems adds another driving force affecting grain growth. In addition, the influence of the heat of mixing (DELTAH(mix)) on atomic mobility and boundary migration was incorporated in the model via a Darken effect.

Experiments were performed to evaluate the effect of 1.5 MeV Kr irradiation on diffusion-induced grain boundary migration (DIGM) in Au/Cu bilayers in the temperature range of 300 ≤ T ≤ 550 K. The experimental results were consistent with DIGM occuring in bilayers both during irradiation and during annealing treatments. Rutherford backscattering spectrometry showed a nearly uniform distribution of Cu present through the entire thickness of appropriately prepared polycrystalline Au films irradiated or annealed at temperatures ≥ 400 K. No parallel effect was seen in similarly treated single-crystal films. In each polycrystalline sample studied, irradiation resulted in greater amounts of Cu present uniformly in the Au compared to annealing-only. The magnitudes of measured Cu compositions were substantially greater than that expected solely from grain boundary diffusion. A simple analysis of the process indicated that ion irradiation affects DIGM by increasing the composition of Cu present in alloyed zones and/or by increasing the grain boundary velocities in the Au.

Diffusion-induced grain boundary migration (DIGM) was observed in Au/Cu bilayers irradiated with 1.5 MeV Kr at T≥400 K. Rutherford backscattering spectrometry (RBS) showed nearly uniform distributions of Cu present throughout polycrystalline Au films after irradiation and after annealing treatments. Irradiation increased the amount of Cu relative to annealed-only areas. Cross-section transmission electron microscopy combined with x-ray energy dispersive spectroscopy (XEDS) identified alloyed zones (14–20 at. % Cu), confirming DIGM in the Au film of an ion bombarded bilayer. A description of DIGM is presented relating RBS measurements of the film-averaged Cu composition with treatment time, average grain size, and film thickness. Application of this model to the experimental results in combination with XEDS work indicates that irradiation enhances DIGM by increasing the grain boundary velocity.

A model of ion-induced grain growth is developed incorporating the irradiation effect concept of thermal spikes. The results of the model predict that for normal grain growth the ion-induced mobility is linearly proportional to the quantity F2D/ΔH3coh, where FD is the ion and recoil energy deposited in nuclear interactions and ΔHcoh is the cohesive energy. This linearity is shown to be supported by the data from six of seven previous ion-induced grain growth experiments. The model analysis is combined with the experimental data to determine values of the proportionality constant, βIIGG, relating the cohesive energy to the activation energy for grain growth (Q> = -βIIGGΔHcoh). The values are found to span a range, 0.05 ≤ βIIGG ≤ 0.10, which is less than the value previously determined for the thermal spike treatment of ion beam mixing (βIM=0.14), and therefore consistent with the idea that atom migration across grain boundaries is easier than migration within the lattice. The consistency of results from the analysis of an entirely different phenomenon adds further credence to the thermal spike treatment of ion-induced grain growth. Finally, it is recommended that additional experiments be performed to evaluate further the model's validity.

Irradiation experiments were conducted on multilayer (ML) and coevaporated (CO) thin films in order to examine the role that the heat-of-mixing (ΔHmix) has in ion-induced grain growth. Room-temperature irradiations using 1.7-MeV Xe ions were performed in the High Voltage Electron Microscope at Argonne National Laboratory. The ML films (Pt-Ti, Pt-V, Pt-Ni, Au-Co, and Ni-Al) spanned a large range of calculated ΔHmix values. Comparison of grain growth rates between ML and CO films of a given alloy confirmed a heat-of-mixing effect. With the exception of the Pt-V system, differences in grain growth rates between ML and CO films varied according to the sign of the calculated ΔHmix of the system. Substantial variations in growth rates among CO alloy films experiencing similar displacement damage demonstrated that a purely collisional approach is inadequate for describing ion-induced grain growth. Therefore consideration must also be given to material-specific properties, such as cohesive energy.

Irradiation experiments were conducted on multilayer (ML) and coevaporated (CO) thin films in order to examine the role that the heat-of-mixing (ΔHmix) has in ion-induced grain growth. Room-temperature irradiations using 1.7 MeV Xe were performed in the High Voltage Electron Microscope at Argonne National Laboratory. The alloys studied (Pt-Ti, Pt-V, Pt-Ni, Au-Co and Ni-Al) spanned a large range of ΔHmix values. Comparison of grain growth rates between ML and CO films of a given alloy confirmed a heat of mixing effect. Differences in grain growth rates between ML and CO films scaled according to the sign and magnitude of ΔHmix of the system (with the exception of the Pt-V system). Substantial variations in growth rates among CO alloy films experiencing similar irradiation damage demonstrated that a purely collisional approach is inadequate for describing ion-induced grain growth and consideration must also be given to material-specific properties. Results from CO alloy films were consistent with a thermal spike model of ion-induced grain growth. The grain boundary mobility was observed to be proportional to the thermal spike-related parameter, F2DΔH3coh, where FD is the energy deposited in nuclear interactions and ΔHcoh is the cohesive energy.

Phase formation was studied in ion-irradiated multilayer and coevaporated Ni-20 at. % Al films supported by Cu, Mo, and Ni transmission electron microscopy (TEM) grids. Irradiation with either 700-keV Xe or 1.7-MeV Xe, to doses sufficient to homogenize the multilayers (≥7.5×1015 cm−2), resulted in the formation of metastable supersaturated γ and HCP phases in both film types. Post-irradiation annealing of multilayers at 450 °C for 1 h transformed the metastable phases to a two-phase γ+γ′ microstructure. In the absence of Cu, the formation of γ′ appeared to proceed by a traditional diffusional growth mechanism, resulting in small (<50 Å) γ′ precipitates in γ matrix grains. The presence of Cu caused the formation of a dual-phase γ+γ′ structure (i.e., distinct, equal-sized grains of γ and γ′) during post-irradiation annealing. It is suggested that copper affected the nucleation of γ′ precipitates and increased the kinetics of growth resulting in the dual-phase morphology. Strong irradiation-induced textures were observed in the multilayers that were less pronounced in the coevaporated films. The texture in the multilayers was attributed to the presence of a slight as-evaporated texture combined with the enhanced atomic mobility due to the heat-of-mixing released during irradiation. The irradiation-induced texture appeared to be necessary for the formation of the dual-phase structure since it likely provided high-diffusivity paths for Cu to diffuse into the film from the TEM grid.

Flexural fatigue tests at surface strain amplitudes of ±0.0012 were conducted on polycrystalline nickel to determine the effect of ion beam surface modifications of the Ni/Al system on fatigue crack initiation. An Ni-75at.%Al surface region 1000 Å thick was produced by evaporating alternate layers of nickel and aluminum followed by ion beam mixing with 3 MeV Ni2+ ions to a fluence of 1 × 1016 ions cm−2. A two-phase γ-γ′ surface structure was produced through implantation of 350 keV Al+ ions to a fluence of 5 × 1017 ions cm−2 into the nickel specimens at 723 K. Nickel specimens that were self-implanted with 350 keV Ni+ ions to a fluence of 1 × 1016 ions cm−2 were also analyzed.