abstract: 2.a
Effects of Non-Glide Stresses on Plastic Flow: from Atomistic Studies of Dislocations to Macroscopic Failure Mechanisms
JOHN L.
BASSANI,
University of Pennsylvania, USA.
In crystals with open structures the cores of dislocations can spread significantly onto several non-parallel planes. The most widely recognized example is the screw dislocation in bcc metals, though this phenomenon is quite ubiquitous in structures that are not close packed including intermetallic compounds. As a consequence, stress components in addition to the Schmid stress affect dislocation mobility. Atomic-level simulations of dislocations are utilized to identify the important non-glide stress components and to construct functional forms of yield criteria entering single crystal constitutive relations. In bcc metals, non-glide components of stress parallel to the Burgers vector as well as shear stresses in 
planes perpendicular to the Burgers vector are found to be important in studies of Mo and Ta.
Based upon these findings, and consistent with experiments, we have developed multiple-slip constitutive relations at the single crystal level that are of the non-associated flow type (behavior which is characteristic, for example, of frictional effects where the normal stress also affects sliding). For the most part, non-associated flow has been ignored in engineering models of plasticity, but its implications are indeed significant. In order to demonstrate the latter, we have studied polycrystalline behavior and failure mechanisms. Taylor averages have been used to construct yield surfaces and flow potentials for both random and textured polycrystals. Non-associated flow is found be present at that scale as well. We have derived simple functions for macroscopic yield surfaces and flow potentials in the case of isotropic behavior and used these to study strain localization and forming limits. The effect of non-associated flow on these failure mechanisms is predicted to be significant. All of this raises interest in another important signature of non-planar core configurations: hypersensitivity to small traces of certain interstitial solutes. Consequently, there is tremendous potential for alloy design of high-strength materials based upon multiscale modeling from quantum mechanics on up.
abstract: 2.b
Stochastic Processes in Collective Dislocation Behaviour
PETER HÄHNER,
European Commission, JRC-IE, The Neterlands.
Research into the microstructural fundamentals of plasticity is driven by the need for an improved understanding of the
interrelation between the composition, processing and structure of materials, their micromechanical behaviour, and their
ultimate macromechanical performance. One of the issues of interest relates to the analytical and numerical modelling of
the dislocation dynamics at the mesoscopic scale, where various spatio-temporal patterning phenomena are induced by the
plastic deformation. The present paper aims at setting out a basic theoretical framework for the description of dislocation
patterning from the viewpoint of self-organization in non-linear dynamic systems. Modelling approaches have to cope with
the long-range nature of dislocation interactions. From a statistical mechanics point of view, a major challenge originates
from the fact that simple coarse graining procedures may not be applicable, since various length scales cannot be clearly
separated owing to the collective dislocation behaviour, which is mediated by the long-range interactions. Adopting a statistical picture
of collective dislocation fluctuations at the mesoscopic scale, a stochastic dislocation dynamics is formulated, which
emphasizes the importance of the strain-rate sensitivity of the flow stress in controlling the effective range of
dislocation interactions. Various applications of the stochastic approach are presented: (i) mesoscopic spatial dislocation
patterning during monotonic multiple-slip deformation (dislocation cell structures), (ii) fatigue dislocation structures
during cyclic single-slip deformation (persistent slip bands), and (iii) macroscopic spatio-temporal strain localization
associated with plastic instability of the strain-rate softening type (Portevin-LeChatelier effect and thermo-mechanical
instabilities). Implications for numerical modelling approaches are also discussed.
abstract: 2.c
Massively Parallel Dislocation Dynamics and Crystal Plasticity
WEI
CAI,
VASILY V.
BULATOV,
TIM G.
PIERCE,
MASATO
HIRATANI,
MOONO
RHEE,
MEIJIE
TANG,
Lawrence Livermore National Laboratory, USA.
Prediction of the plastic strength of single crystals based on the collective dynamics of dislocations has been a grand challenge for computational materials science for a number of years. The difficulty lies in the inability of the available dislocation dynamics (DD) codes to handle a sufficiently large number of dislocation lines, in order to be statistically representative and to reproduce experimentally observed microstructures. Our new massively parallel DD code is capable of modeling million dislocations simultaneously by employing thousands of processors. We will discuss the methods that make simulations of such scale possible. Simulation data on strain hardening and spontaneous dislocation patterning in FCC and BCC metals will be presented and compared with experiments. We will discuss what sort of new information can be extracted from such simulations and examine how close we are able to come to understanding single crystal plasticity from the underlying collective motion of dislocations.
abstract: 2.d
Disorientations in Dislocation Structures
W.
PANTLEON,
Risoe National Laboratory, Roskilde, Denmark.
Probably the most striking feature of dislocation ensembles is their (self-)organization into hierarchical dislocation structures. Most of the stored dislocations gather in dislocation boundaries and different boundary types can be identified with respect to their morphology. Beside the width of the boundaries and their spacing, they are characterized by the orientation difference between the adjacent dislocation-free regions. The disorientations are a direct consequence of non-redundant dislocations, i.e. excess dislocations of one sign of a Burgers vector - in contrast to redundant dislocations, which do not have any geometrical effect on this length scale. The interest in disorientations in deformation structures has increased enormously since automatic orientation measurements have become available and orientation imaging maps are easily obtained. The experimental knowledge based on TEM and EBSD is summarized and some of the controversial issues discussed.
Based on dislocation dynamics the emergence of disorientations with plastic deformation is modelled. Incorporating different storage mechanisms for dislocations, the model predicts the disorientation angles (average and distribution) across different types of boundaries successfully and in close agreement with experimental data. Spatial correlations between the orientation differences in neighbouring boundaries are also traced to the accumulation mechanism. This allows distinction between alternating and random disorientations - the later are sometimes mistaken as orientation gradients. Evidences for spatial correlations can be gained from orientation imaging maps, but also from X-ray diffraction rocking curves.
Additionally, the evolution of the total dislocation density from dislocation model is utilized for predicting the resulting flow stress. Due to the same accumulation and annihilation mechanisms as in Kocks' model a stage III behaviour is obtained. This behaviour is not changed (at least qualitatively) as long as disorientations arise from statistical reasons only. If there are any additional deterministic contributions to dislocation accumulation as either in geometrically necessary boundaries or in orientation gradients, an extra work-hardening term arises from a reduced annihilation of excess dislocations caused by the lack of annihilation partners. The resulting constant work-hardening rate is characteristic for stage IV and in good accordance with experimental findings.