Extensive simulations of interactions between moving dislocations and dislocation-like obstacles such as DLs and SFTs has demonstrated that the reactions involved follow the general rules of dislocation — dislocation reaction, for example, Frank’s rule for Burgers vectors,1,2 even though the reacting segments are of the nanometer scale in length. Results of these interactions are in the range from no effect on both dislocation and obstacle to complete disappearance of the obstacle and significant modification of the dislo­cation. A detailed analysis of reactions was made for SFTs an fcc metal56 and later for SIA loops in Fe.57 In general, five types of reaction were identified, as sum­marized in Table 1. The outcomes in Table 1 were observed for different obstacles under different reac­tion conditions such as interaction geometry, strain rate, ambient temperature, and so on. We give some examples in the following section.

1.12.4.2.1 Stacking fault tetrahedra

Reactions of type R1 have been observed for both screw and edge dislocations and all the defects with dislocation character. Interestingly, the strength effect of this reaction varies from minimum to

maximum. For example, it is insignificant in the case of a 1/2(110) {111} edge dislocation interacting with an SFT58 and maximum for a screw dislocation inter­acting with a DL when the loop is fully absorbed into a helical turn on the dislocation.57 The mechanism for the way both the obstacle and dislocation remain unchanged is different for each case. An edge dislo­cation interacting with an SFT close to its tip creates a pair of ledges on its surface that are not stable and annihilate athermally.56,58 An example of this reaction is presented in Figure 10. If the dislocation slip plane is far enough from the SFT tip in the compressive region of the dislocation (for details of geometrical definitions see Bacon et a/.4), the ledges can be stabilized.56,58,59 This can be seen in Figure 11 (1). If the dislocation passes through the SFT several times in the same slip plane, it can detach the portion of the SFT above the slip plane, as shown in Figure 11 (2-4). Both the above mechan­isms are common for small SFTs, low T, fast disloca­tions, and the position of the SFT tip above the slip plane of an edge dislocation. If, however, the SFT tip is below the dislocation slip plane, and T is high enough and the dislocation speed low enough,

reaction R3 can be activated. The stages of this reac­tion are presented in Figure 12. An example of effects of SFT orientation and temperature for the interaction of an edge dislocation with an SFT is presented in Figure 13. In this study, the dislocation slip plane intersected a 4.2 nm SFT through its geo­metrical center at the applied є = 5 x 106s-1 in a wide temperature range from 0 to 450 K.59 Reaction R1 was observed (see Figure 10) at all temperatures when the SFT was oriented with its tip up relative to the dislocation slip plane (orange triangles up in Figure 13) and at the two lowest temperatures when it was oriented in the opposite sense. At T = 300 K and orientation with tip down, a couple of ledges were formed on the SFT surface (see Figure 11 ). It may be noted that the R2 mechanism requires higher applied stress even though the tem­perature is increased. At higher T = 450 K, the inter­action mechanism is changed and the whole portion of the SFT above the slip plane is absorbed by the dislocation (Figure 12), that is, reaction R3 occurs, creating a pair of superjogs on the dislocation line. Some vacancies were also found to form to accom­modate the glissile configuration of the superjogs.

This is discussed later in Section 1.12.4.3. More details on interactions between screw and edge dis­locations and SFT can be found elsewhere.60-63

In general, it can be concluded that the SFTs created under irradiation, that is, <4nm in size,6 are very stable objects and unlikely to be eliminated by a simple interaction with either edge or screw dislocations. Numerous attempts have been made to find a mechanism responsible for formation of clear, defect-free channels in irradiated fcc materials.64 One of the most used models considers absorption of an SFT by screw and mixed 60° dislocations.65 The absorption by conversion of an SFT into a heli­cal turn on a screw dislocation has been observed by in situ transmission electron microscope (TEM) deformation experiments,64,66-68 but only partial absorption has been found in MD modeling. As observed and discussed elsewhere,4,59 SFT separation into parts due to temporary absorption of part of an SFT as a helical turn and its expansion along a screw dislocation line can occur, but complete annihilation of an SFT has not been reproduced by atomic-scale
modeling of bulk material. Possible reasons, including inability of MD to reproduce the whole set of exper­imental conditions such as stress state, scale, strain rate, and so on, were discussed in Matsukawa eta/.66-68 An alternative interpretation was suggested as a result of MD simulation of interaction between a screw dislocation and an SFT in a thin film,26 that is, the conditions realized experimentally for in situ TEM deformation. A thin film of fcc Cu was simulated and a 1/2(110) screw dislocation with an end on each surface was moved toward the SFT (size 12 and 18 nm) placed in the film center. A number of steps occurred that resulted in elimination of the vast portion of the SFT26:

1. The dislocation glided toward the SFT and par­tially absorbed it as a helical turn.

2. Edge segments of the turn glided toward the free surfaces and were annihilated there.

3. Glide of edge segments provided mass transport; in this particular case, transport of vacancies to the free surfaces.

4.
The screw character of the dislocation was restored but in another slip plane. The dislocation then glided away leaving behind a small portion of the original SFT.

Although this mechanism provides a mechanistic understanding of in situ TEM observations, it can operate only between surfaces or interfaces where the ends of the screw dislocation can cross-slip and, therefore, cannot be applied to bulk material and is unlikely to be responsible for the clear channel for­mation observed in bulk samples. We will return to this question later.

1.12.4.2.2 Dislocation loops

DLs are common objects formed in metals under irra­diation. Depending on metal properties and irradiation conditions, loops ofdifferent types can be formed. They are mainly interstitial in nature although vacancy loops can appear in some metals under specific conditions. Here we consider only interstitial DLs. The shortest Burgers vector in bcc metals is bL = 1/2(111) and loops with this are the most common. In neutron and heavy-ion irradiated Fe, loops can have bL = (100),69 particularly at Tabove about 300 °C. Note that both have a perfect Burgers vector and are glissile. The most common loops in fcc metals have bL = 1/2(110) and are also perfect and glissile. In metals and alloys with a low stacking fault energy, Frank loops with bL = 1/3(111) can form. They are faulted and sessile. In the following section, we consider examples of reactions R1-R4 in fcc and bcc metals.