Yu.N. Osetsky

The primary damage state in fcc, bcc and hcp metals as seen in molecular dynamics simulations

Recent progress in the use of molecular dynamics (MD) to investigate the primary state of damage due to displacement cascades in metals is reviewed, with particular emphasis on the influence of crystal structure. Topics considered include the effect on defect formation in pure metals and alloys of primary knock-on atom (PKA) energy and irradiation temperature. An earlier empirical relationship between the production efficiency of Frenkel pairs and cascade energy is seen to have wide validity, and the reduction in efficiency with increasing irradiation temperature is small. Crystal structure has little effect on the defect number. In terms of the development of models to describe the evolution of radiation damage and its role in irradiation-induced changes in material properties, the important parameters are not only the total number of Frenkel defects per cascade but also the distribution of their population in clusters and the form and mobility of these clusters. Self-interstitial atoms form clusters in the cascade process in all metals, and the extent of this clustering does appear to vary from metal to metal. Vacancy clustering is also variable. The mobility of all clusters depends on their dislocation character and thus on the crystal structure and stacking fault energy. It is shown that computer simulation can provide detailed information on the properties of these defects.

One-dimensional atomic transport by clusters of self-interstitial atoms in iron and copper

Atomic-scale computer simulation has been used to study the thermally activated atomic transport of self-interstitial atoms (SIAs) in the form of planar clusters in pure Cu and f-Fe. There is strong evidence that such clusters are commonly formed in metals during irradiation with high-energy particles and play an important role in accumulation and spatial distribution of surviving defects. An extensive study of the mobility of SIA clusters containing two to 331 interstitials has been carried out using the molecular dynamics simulation technique for the temperature range from 180 to 1200 K. The results obtained show that clusters larger than three to four SIAs are one-dimensionally mobile in both Cu and Fe. Large clusters of more than 100 SIAs in Cu and 300 SIAs in Fe have significantly reduced mobility. The problem of describing one-dimensional (1D) motion in three-dimensional space is discussed. An attempt is made to describe the mobility of SIA clusters within the approximation of 1D diffusion. For clusters in both metals the effective migration energy of 1D diffusion as estimated via the jump frequency of the cluster centre of mass is found to be independent of the number of SIAs in the clusters, although the cluster jump frequency decreases with increasing cluster size. Mechanisms of 1D mobility of interstitial clusters are discussed.

Aspects of microstructure evolution under cascade damage conditions

Abstract The conventional theoretical models describing the damage accumulation, particularly void swelling, under cascade damage conditions do not include treatments of important features such as intracascade clustering of self-interstitial atoms (SIAs) and one-dimensional glide of SIA clusters produced in the cascades. Recently, it has been suggested that the problem can be treated in terms of ‘production bias’ and one-dimensional glide of small SIA clusters. In the earlier treatments a ‘mean size approximation’ was used for the defect clusters and cavities evolving during irradiation. In the present work, we use the ‘size distribution function’ to determine the dose dependence of sink strengths, vacancy supersaturation and void swelling as a function of dislocation density and grain size within the framework of production bias model and glide of small SIA clusters. In this work, the role of the sessile-glissile loop transformation (due to vacancy supersaturation) on the damage accumulation behaviour is included. The calculated results on void swelling are compared with the experimental results as well as the results of the earlier calculations using the ‘mean size approximation’. The calculated results agree very well with the experimental results.

Atomic modelling of strengthening mechanisms due to voids and copper precipitates in α-iron

Recently a model has been developed by Osetsky and Bacon to study edge dislocations moving over large distances on the atomic scale. It permits investigation of motion of a dislocation under different conditions of applied shear stress with constant or variable strain rate and temperature, and in the presence of obstacles. In this paper we apply the model to study the motion of an infinite straight but flexible edge dislocation through a row of either voids or coherent copper precipitates in bcc iron. Stress–strain curves, energy barrier profile and strength characteristics of obstacles and other dislocation configuration information have been obtained from the modelling and compared with continuum treatments. Some specific atomic-scale mechanisms associated with strengthening due to voids and precipitates over a range of size have been observed and discussed.

Structure and properties of clusters of self-interstitial atoms in fcc copper and bcc iron

Abstract Static and molecular dynamics simulations have been used with different types of interatomic potentials to investigate the structure, properties and stability of self-interstitial atom (SIA) clusters produced during irradiation. In α-iron (Fe), faulted clusters of <110> dumbbells are unstable for all the potentials. The most stable SIA clusters are sets of parallel <111> crowdions. Large clusters of this type form perfect dislocation loops with Burgers vector b = ½⟨111⟩. Small clusters (less than 9 SIAs) of ⟨100⟩ crowdions are stable at 0K, but transform into a set of ⟨111⟩ crowdions on annealing. Larger ⟨100⟩ clusters are stable and form perfect dislocation loops with b = ⟨100⟩. Both types of loops are glissile. In copper (Cu), clusters of parallel ⟨100⟩ dumbbells and ⟨110⟩ crowdions are stable. Large clusters of these types form faulted and perfect dislocation loops with b = ⅓ ⟨111⟩ and ½ ⟨110⟩ respectively. Small faulted clusters (less than 7 SIAs) of irregular shape can transform into a set of ⟨110⟩ crowdions during annealing. Larger faulted clusters are stable as hexagonal ⅓ ⟨111⟩ Frank loops at temperatures of about up to 1050K for a period of several hundred picoseconds. All faulted clusters are sessile. Clusters of ⟨110⟩ crowdions and ½ ⟨110⟩ perfect loops are glissile and stable at all temperatures. When large enough (more than 49–64 SIAs) they can dissociate on their glide prism. Symmetric three-dimensional clusters of ⟨100⟩ dumbbells are stable at 0K but during annealing they transform into sets of ⟨110⟩ crowdions. The results for both iron and copper are discussed and compared with experimental data and provide a basis for investigating and explaining the observed differences in radiation damage accumulation behaviour between fcc and bcc metals.

Atomistic study of the generation, interaction, accumulation and annihilation of cascade-induced defect clusters

Recent theoretical calculations and atomistic computer simulations have shown that glissile clusters of self-interstitial atoms (SIAs) play an important role in the evolution of microstructure in metals and alloys under cascade damage conditions. Over the past decade or so, the properties of SIA clusters in fcc, bcc and hcp lattices have been widely studied. In this paper we review key properties of these defects and also those of vacancy clusters formed directly in cascades, and present an atomic-level picture based on computer modelling of how these properties may change in the presence of other defects, impurities, stress fields, etc. We then examine the role of cluster properties and the consequences of their interactions in the process of damage accumulation and changes in mechanical and physical properties. We focus on the formation of defect clusters (e.g. dislocation loops and stacking fault tetrahedra (SFT)) and their segregation in the form of rafts of dislocation loops and atmospheres of loops decorating dislocations. Finally, we address the problem of radiation hardening by considering interactions between mobile dislocations and defect clusters (e.g. SIA dislocation loops, SFT and microvoids) produced during irradiation.

Thermally activated glide of small dislocation loops in metals

Recent computer simulation studies have shown that small clusters of selfinterstitial atoms (SIAs) formed in displacement cascades consist of crowdions and are highly mobile in the crowdion direction. In the present work, we use molecular dynamics to investigate whether small perfect vacancy loops formed in cascades in alpha-Fe and Cu are also mobile. Loops containing more than about 30 vacancies in Fe are found to produce atomic displacements during annealing, due to thermally activated movement in the direction of their Burgers vector that is qualitatively similar to the mechanism of SIA cluster motion. Although vacancy clusters are slightly less mobile than SIA clusters under the same conditions, their mobility is significantly higher than that of the monovacancy. The motion of vacancy loops in Cu does not occur because they transform into sessile configurations similar to stacking-fault tetrahedra. These results point to the possibly important contribution of vacancy loop mobility to the difference in...

Defect cluster formation in displacement cascades in copper

Extensive study of primary damage in displacement cascades in metals by computer simulation has shown that the total number of defects produced is significantly lower than predicted by the Norgett, Robinson and Torrens (NRT) model and that a significant fraction of the self-interstitials forms glissile clusters. However, there is a lack of variety of defect types observed in cascade simulation, which, in many cases, makes it difficult to explain experimental data. For example, experiments on copper show efficient production of stacking fault tetrahedra (SFTs) but they were not observed systematically in computer simulation. To consider this further, extensive simulation of displacement cascades in copper has been performed using two different interatomic potentials, a short-range many-body potential (MBP) and a long-range pair potential (PP). Primary knock-on-atom (PKA) energy in the range 2–20 keV and temperatures of 100 and 600 K were considered. Special attention was paid to cascade statistics and the accuracy of simulation in the collision stage. The former required many simulations for each energy whereas the latter involved a modification of the simulation method to treat a hot region with high accuracy by applying a smaller time step. Results showing the variety of clusters observed, e.g. SFTs, glissile and sessile interstitial clusters, and faulted and perfect interstitial dislocation loops, are presented.

Interactions between mobile dislocation loops in Cu and α-Fe

The relevance of one-dimensionally gliding clusters in the understanding of the damage accumulation produced by the displacement cascades has been underlined by the production bias model. The properties and mobility of isolated clusters of vacancies and self-interstitials have been recently studied by molecular dynamics and valuable information about their diffusional characteristics is obtained. The next step in the understanding of radiation damage should include the possible reactions of these clusters with other clusters, dislocations and other sinks. In this paper we present the first results of a molecular dynamics study of interactions between glissile interstitial clusters and small dislocation loops in α-Fe and Cu. Different types of interactions have been studied between clusters of different sizes (from 12 to 91 defects) in the temperature range from 300 to 1000 K. As a result of the inter-cluster interactions both glissile and sessile clusters can be obtained and this depends on the metal, reaction type and size of the clusters. In general the probability to form sessile clusters increases for larger clusters and it is higher in Cu. The results obtained are discussed from the point of view of the difference in radiation damage effects in fcc Cu and bcc Fe.

Computer simulation of displacement cascades and the defects they generate in metals

Abstract Molecular dynamics (MD) computer simulation provides information that cannot be obtained by other means on the primary state of damage due to displacement cascades in metals. Progress in a number of topics in this field is reviewed here. It includes research dealing with the effect on defect formation in pure metals and alloys of primary knock-on atom (PKA) energy and irradiation temperature. Clear views on dependencies and trends have emerged in these areas. In terms of the development of models to describe the evolution of radiation damage and its role in phenomena such as irradiation-induced hardening, creep and swelling, the important parameters are not only the total number of Frenkel defects but also the distribution of their population in clusters and the form and mobility of these clusters. Recent results on these aspects are presented and it is shown that computer simulation provides detailed information that paves the way for successful development of models of the evolution of damage beyond the stage of the cascade process.