 |
Center for Nonlinear Studies |
Modeling plastic deformation is an inherently multiscale problem. While much work has been done at understanding the atomic-level foundations of plasticity in a wide variety of materials, it is not straightforward to utilize these results to make macroscopic predictions. One of the major obstacles is to develop a systematic multiscale framework in which the relevant details obtained at lower scalespercolate systematically into mesoscopic and continuum models. In the past 10 years, we have been working on developing this framework for the description of plasticity in materials that do not obey the Schmid law. These include all body-centered cubic (bcc) metals of the VB and VIB groups, Fe, alkali metals, and the hexagonal crystals deforming by non-basal slip. Focusing on bcc metals Mo and W, I will explain why first principles calculations are needed to obtain accurate descriptions of bonding in these materials. This will be employed to obtain an atomistic database that describes under which conditions the screw dislocations in these materials become mobile. These results are used to formulate a single crystal yield criterion for bcc metals that is shown to be very different from those due to von Mises or Tresca. To extend these results to finite temperatures and strain rates, we developed a thermodynamic model of slip that depends on the shape of the Peielrs barrier and its changes understress. These can be calculated efficiently by our recent generalization of the Nudged Elastic Band method that takes into account atomic relaxations. The activation enthalpies thus obtained are used to develop a discrete dislocation dynamics model and the mean field model of plasticity. These results have been implemented into crystal plasticity finite element models for single crystals and random polycrystals that can now be used to make quantitative predictions. Finally, I will outline a series of experiments that are now under way to validate this multiscale framework.
Host: Turab Lookman