While atomistic methods can probe the primary damage state with great detail, they can also be used to probe the interactions of the defects formed with the underlying microstructure. An example is the case of creep due to irradiation in materials. Creep of metals and alloys under irradiation has been the subject of many experimental and theoretical studies for more than 30 years. Although a vast amount of knowledge of irradiation creep has accumulated, the database on irradiation creep comes from many relatively small experiments, and there were often differences in experimental conditions from one study to the next. Theoretical models are based on linear elasticity. Among the many theories that exist to describe the driving force for irradiation creep, the most important are the SIPN, SIPA, and SIPA-AD effects.

Stress Induced Preferential Nucleation of loops (SIPN) is based on the idea that the application of external stress will result in an increased number of dislocation loops nucleating on planes of preferred orientations. Interstitial loops will tend to be oriented perpendicular to the applied tensile stress, while vacancy loops will prefer to be oriented parallel to the stress. The net result is elongation of the solid in the direction of applied stress. While there is some experimental support of this theory, it is thought that it cannot account fully for creep seen in materials.

An alternative theory is Stress Induced Preferential Absorption/Attraction (SIPA). The essential idea behind SIPA is that interstitials are preferentially absorbed by dislocations of particular orientations, resulting in climb; this is described by an elastic interaction between the stress fields of the defect and dislocation. A variant on SIPA that accounts for anistropic diffusion is SIPA-AD. This theory uses the full diffusion equations, derived by Dederichs and Schroeder (1978), to take into account anisotropic stress fields. Savino and Tome developed this theory and found that it generally gives a larger contribution to dislocation climb than the original SIPA (Tome, Cecatto et al.). A thorough review of many dislocation creep models was prepared by Matthews and Finnis (1988).

These models go a long way towards explaining irradiation creep due to dislocations. However, all models based on linear elasticity break down near a dislocation core due to the 1/r terms in the stress and strain field expressions. Atomistic calculations do not suffer from this problem, so they can be used to verify the range of validity of theoretical expressions and successfully predict true behavior at the core.

Molecular statics calculations can be performed in order to understand the interactions between vacancies and interstitials and line dislocations in bcc iron. These can be compared to similar results given by dipole tensor calculations based in linear elasticity theory. Results from two methods are used to calculate the interaction energy between a dislocation and a point defects in bcc iron are compared. For vacancies and a variety of self-interstitial dumbbell configurations near both edge and screw dislocation cores, there are significant differences between direct calculations and atomistics. For vacancies some interaction is seen with both edge and screw dislocations where none is predicted. Results for interstitials tended to have a strong dependence on orientation and position about the core. Particularly for the screw, continuum theory misses the tri-fold splitting of the dislocation core which has a large influence on atomistic results.

Fig. 4d. Final position of the dislocation with the interstitial absorbed in the core

Figures 4a-d shows the evolution of a defect in the vicinity of the dislocation. In this example an interstitial is positioned near a dislocation core and the energy of the system is minimized. The interstitial moves into the dislocation core and forms an extended jog in the dislocation core structure. Such a process is not captured by linear elasticity calculations which fail to capture the core structure and the core-defect interactions.

The relationship between crystal plasticity and dislocation behavior in materials has motivated a wide range of experimental and computational studies of dislocation behavior (Vitek 1976; Osetsky, Bacon et al. 1999). Computational studies of dislocation activity can be performed at several different length and time scales. In some cases, a multi-scale modeling approach is adopted(Ghoniem et al. 2003). Core properties and atomic mechanisms are simulated using first principles calculations and molecular dynamics simulations. These results can be used to form the rules that govern large-scale Dislocation Dynamics (DD) simulations (Wen et al. 2005) that account for the activity of a large number of dislocation segments. Polycrystalline plasticity models are then developed that utilize the information at the atomistic scale to parameterize partial differential equations of rate dependent viscoplasticity(Deo, Tom et al. 2008). While understanding dislocation creep processes, dislocation climb rates and hence, the interaction of dislocations with point defects is an important quantity to be calculated. Here, we show how the dislocation core affects the interaction energy between the dislocation and the point defect using both linear elasticity as well as atomistic calculations.